tempest-react-sdk 0.69.0 → 0.71.0

package/dist/vision/tasks/classifier.js

@@ -1,17 +1,18 @@
 import { SpeedTimer as e } from "../core/timing.js";
 import { ClassificationResults as t, Probs as n } from "../results.js";
 import { resolveLabels as r } from "../labels.js";
-import { classificationNumClasses as i, resolveInputSize as a } from "../core/graph.js";
-import { modelNames as o } from "../core/metadata.js";
-import { OrtSession as s } from "../core/session.js";
-import { isUltralyticsClassifier as c, resolveNormalization as l } from "../normalization.js";
-import { loadImage as u } from "../io/image.js";
-import { toFloat32Tensor as d } from "../preprocess/image.js";
-import { ResizePipeline as f, zeroTensorData as p } from "../preprocess/pipeline.js";
-import { softmax as m, topK as h } from "../postprocess/classification.js";
-import { VisionTask as g } from "./base.js";
+import { asFloat32Array as i } from "../core/dtypes.js";
+import { classificationNumClasses as a, resolveInputSize as o } from "../core/graph.js";
+import { modelNames as s } from "../core/metadata.js";
+import { OrtSession as c } from "../core/session.js";
+import { isUltralyticsClassifier as l, resolveNormalization as u } from "../normalization.js";
+import { loadImageSource as d } from "../io/image.js";
+import { toFloat32Tensor as f } from "../preprocess/image.js";
+import { ResizePipeline as p, zeroTensorData as m } from "../preprocess/pipeline.js";
+import { softmax as h, topK as g } from "../postprocess/classification.js";
+import { VisionTask as _ } from "./base.js";
 //#region src/vision/tasks/classifier.ts
-var _ = class _ extends g {
+var v = class v extends _ {
 	_labels;
 	_names;
 	_inputSize;
@@ -32,7 +33,7 @@ var _ = class _ extends g {
 	async warmup(e = 1) {
 		let [t, n] = this._inputSize;
 		for (let r = 0; r < e; r++) {
-			let e = d(p(t, n), [
+			let e = f(m(t, n), [
 				1,
 				3,
 				n,
@@ -44,22 +45,22 @@ var _ = class _ extends g {
 	get _pipeline() {
 		if (this._pipelineCache === null) {
 			let [e, t] = this._inputSize;
-			this._pipelineCache = new f(e, t, this._mean, this._std);
+			this._pipelineCache = new p(e, t, this._mean, this._std);
 		}
 		return this._pipelineCache;
 	}
 	static async create(e, t = {}) {
-		let n = await s.create(e, t), u = l(n.metadata, {
+		let n = await c.create(e, t), i = u(n.metadata, {
 			normalization: t.normalization,
 			mean: t.mean,
 			std: t.std
-		}), d = t.numClasses ?? i(n.outputShape) ?? void 0, f = r(t.labels ?? o(n.metadata), { numClasses: d }), p = {};
+		}), d = t.numClasses ?? a(n.outputShape) ?? void 0, f = r(t.labels ?? s(n.metadata), { numClasses: d }), p = {};
 		for (let e = 0; e < f.length; e++) p[e] = f[e];
-		return new _(n, f, p, a({
+		return new v(n, f, p, o({
 			graphShape: n.inputShape,
 			requested: t.inputSize,
 			fallback: [224, 224]
-		}), u.mean, u.std, t.applySoftmax ?? !c(n.metadata), u.name);
+		}), i.mean, i.std, t.applySoftmax ?? !l(n.metadata), i.name);
 	}
 	get labels() {
 		return this._labels;
@@ -76,56 +77,64 @@ var _ = class _ extends g {
 	async call(e, t = {}) {
 		return this.predict(e, t);
 	}
-	async predict(r, i = {}) {
-		let a = new e(), o = typeof r == "string" ? r : null, s = await u(r);
-		a.stage("load");
-		let c = this._preprocess(s);
-		a.stage("preprocess");
-		let l = await this._session.run({ [this._session.inputName]: c });
-		this._pipeline.release(), a.stage("inference");
-		let d = this._session.outputNames[0];
-		if (d === void 0) throw Error("Classifier model has no outputs.");
-		let f = l[d];
-		if (f === void 0) throw Error(`Classifier model output ${d} missing from run() result.`);
-		let p = this._postprocess(f.data), { indices: m, values: g } = h(p, i.topK ?? null), _ = [];
-		for (let e = 0; e < m.length; e++) {
-			let t = m[e], n = this._labels[t] ?? `class_${t}`;
-			_.push({
+	async predict(r, a = {}) {
+		let o = new e(), s = typeof r == "string" ? r : null, { image: c, canvas: l } = await d(r);
+		o.stage("load");
+		let { tensor: u, reused: f } = this._preprocess(c, l);
+		o.stage("preprocess");
+		let p;
+		try {
+			p = await this._session.run({ [this._session.inputName]: u });
+		} finally {
+			this._pipeline.release({ reused: f });
+		}
+		o.stage("inference");
+		let m = this._session.outputNames[0];
+		if (m === void 0) throw Error("Classifier model has no outputs.");
+		let h = p[m];
+		if (h === void 0) throw Error(`Classifier model output ${m} missing from run() result.`);
+		let _ = this._postprocess(i(h.data)), { indices: v, values: y } = g(_, a.topK ?? null), b = [];
+		for (let e = 0; e < v.length; e++) {
+			let t = v[e], n = this._labels[t] ?? `class_${t}`;
+			b.push({
 				classId: t,
 				className: n,
-				probability: g[e],
+				probability: y[e],
 				cls: t,
 				name: n,
-				conf: g[e]
+				conf: y[e]
 			});
 		}
-		if (_.length === 0) throw Error("Classifier produced no probabilities (empty output).");
-		let v = _[0], y = {
-			classId: v.classId,
-			className: v.className,
-			confidence: v.probability,
-			cls: v.classId,
-			name: v.className,
-			conf: v.probability,
-			image: s,
-			probabilities: _
-		}, b = [s.height, s.width], x = new n(p);
-		return a.stage("postprocess"), [new t(x, y, this._names, s, b, o, a.speed())];
+		if (b.length === 0) throw Error("Classifier produced no probabilities (empty output).");
+		let x = b[0], S = {
+			classId: x.classId,
+			className: x.className,
+			confidence: x.probability,
+			cls: x.classId,
+			name: x.className,
+			conf: x.probability,
+			image: c,
+			probabilities: b
+		}, C = [c.height, c.width], w = new n(_);
+		return o.stage("postprocess"), [new t(w, S, this._names, c, C, s, o.speed())];
 	}
-	_preprocess(e) {
-		let [t, n] = this._inputSize, { data: r } = this._pipeline.run(e);
-		return d(r, [
-			1,
-			3,
-			n,
-			t
-		]);
+	_preprocess(e, t) {
+		let [n, r] = this._inputSize, { data: i, reused: a } = this._pipeline.run(e, t);
+		return {
+			tensor: f(i, [
+				1,
+				3,
+				r,
+				n
+			]),
+			reused: a
+		};
 	}
 	_postprocess(e) {
-		return this._applySoftmax ? m(e) : new Float32Array(e);
+		return this._applySoftmax ? h(e) : new Float32Array(e);
 	}
 };
 //#endregion
-export { _ as Classifier };
+export { v as Classifier };
 
 //# sourceMappingURL=classifier.js.map

package/dist/vision/tasks/classifier.js.map

@@ -1 +1 @@
-{"version":3,"file":"classifier.js","names":[],"sources":["../../../src/vision/tasks/classifier.ts"],"sourcesContent":["/** @generated Vendored from @mauriciobenjamin700/ort-vision-sdk-web. Do not hand-edit — regenerate with `npm run vendor:vision`. */\n/**\n * Image classification task using ONNX Runtime Web.\n */\n\nimport type * as ort from \"onnxruntime-web\";\n\nimport { type ModelSource, type OrtSessionOptions, OrtSession } from \"../core/session\";\nimport { SpeedTimer } from \"../core/timing\";\nimport { type ImageInput, loadImage } from \"../io/image\";\nimport { classificationNumClasses, resolveInputSize } from \"../core/graph\";\nimport { modelNames } from \"../core/metadata\";\nimport { type LabelSpec, resolveLabels } from \"../labels\";\nimport {\n    type Normalization,\n    isUltralyticsClassifier,\n    resolveNormalization,\n} from \"../normalization\";\nimport { softmax, topK } from \"../postprocess/classification\";\nimport { toFloat32Tensor } from \"../preprocess/image\";\nimport { ResizePipeline, zeroTensorData } from \"../preprocess/pipeline\";\nimport { ClassificationResults, Probs } from \"../results\";\nimport { VisionTask } from \"./base\";\nimport { type ClassProbability, type ClassificationResult, type RGBImage } from \"../types\";\n\nexport interface ClassifierOptions extends OrtSessionOptions {\n    /**\n     * Class label spec — see {@link resolveLabels}.\n     *\n     * Optional: when omitted, the names the export baked into the model are used\n     * (Ultralytics writes them as `names` in the metadata map). Only when the\n     * model carries none does this fall back to generated `class_<id>` labels.\n     * Passing a spec always wins, for a model whose names are wrong or absent.\n     */\n    readonly labels?: LabelSpec;\n    /**\n     * Number of classes the model can predict.\n     *\n     * Optional: inferred from the classification head's declared output shape\n     * `(B, nc)`. Pass it to validate that the supplied labels match the model.\n     */\n    readonly numClasses?: number;\n    /**\n     * Model input `[width, height]` in pixels.\n     *\n     * Only used when the model's graph leaves its spatial axes dynamic: a graph\n     * that declares a static size always wins, since that is the only shape ONNX\n     * Runtime will accept. Defaults to `[224, 224]`.\n     */\n    readonly inputSize?: readonly [number, number];\n    /**\n     * Which preprocessing this model expects — see {@link Normalization}.\n     *\n     * Defaults to `\"auto\"`, which reads the model's own export metadata and picks\n     * `\"ultralytics\"` (raw `[0, 1]`) for an Ultralytics classification head and\n     * `\"imagenet\"` for everything else.\n     */\n    readonly normalization?: Normalization;\n    /** Per-channel RGB mean, overriding the preset. Defaults to the preset's. */\n    readonly mean?: readonly [number, number, number];\n    /** Per-channel RGB standard deviation, overriding the preset. */\n    readonly std?: readonly [number, number, number];\n    /**\n     * Whether the model's output still needs a softmax.\n     *\n     * Left undefined (the default), this reads the model's metadata and answers\n     * `false` for an Ultralytics classification export, whose graph already ends\n     * in one — applying a second softmax to a probability vector keeps the\n     * ranking but flattens the confidences, so the top-1 stays right while every\n     * number attached to it is wrong. Detection covers that family; for any other\n     * model that already emits probabilities, pass `false` explicitly.\n     */\n    readonly applySoftmax?: boolean;\n}\n\nexport interface ClassifierPredictOptions {\n    /**\n     * If set, the per-class probability list in `results[0].result.probabilities`\n     * is truncated to the top-K entries. The bulk `probs` view always exposes\n     * the full vector.\n     */\n    readonly topK?: number;\n}\n\n/**\n * Image classifier wrapping an ONNX model with ImageNet-style preprocessing.\n *\n * `predict()` returns `Promise<ClassificationResults[]>` (length 1 for a\n * single image), mirroring Ultralytics' API. The envelope exposes a `probs`\n * collection (`top1`, `top1conf`, `top5`, `top5conf`, `data`) plus the\n * legacy per-class probability list with names already resolved.\n *\n * Defaults: 224×224 RGB input, `float32` normalized with ImageNet mean/std,\n * NCHW layout, batch size 1, softmax applied to the raw output.\n *\n * @example\n * ```typescript\n * const clf = await Classifier.create(\"/models/resnet50.onnx\", {\n *   labels: [\"tench\", \"goldfish\", ...]  // 1000 ImageNet labels\n * });\n * const r = (await clf.predict(\"/images/dog.jpg\"))[0];\n * console.log(r.cls, r.conf, r.name);\n * console.log(r.probs.top5, r.probs.top5conf);\n * ```\n */\nexport class Classifier extends VisionTask {\n    private constructor(\n        session: OrtSession,\n        private readonly _labels: readonly string[],\n        private readonly _names: Readonly<Record<number, string>>,\n        private readonly _inputSize: readonly [number, number],\n        private readonly _mean: readonly [number, number, number],\n        private readonly _std: readonly [number, number, number],\n        private readonly _applySoftmax: boolean,\n        private readonly _normalization: string,\n    ) {\n        super(session);\n    }\n\n    /**\n     * Whether a softmax is applied to the model's output before ranking.\n     *\n     * Resolved once at construction. Worth reading when confidences look\n     * compressed: a second softmax over an already-normalized vector leaves the\n     * ordering intact and the numbers meaningless, which is invisible to any check\n     * that only looks at the predicted class.\n     */\n    get appliesSoftmax(): boolean {\n        return this._applySoftmax;\n    }\n\n    /**\n     * Which preprocessing this classifier applies to every image.\n     *\n     * One of the {@link Normalization} preset names, or `\"custom\"` when the caller\n     * supplied `mean`/`std` directly. Worth reading when a model underperforms:\n     * feeding a classifier a differently prepared tensor than it was trained on\n     * degrades it without throwing anything, so \"what does this assume\" is the\n     * first question.\n     */\n    get normalization(): string {\n        return this._normalization;\n    }\n\n    private _pipelineCache: ResizePipeline | null = null;\n\n    /**\n     * Run the model once on a zero tensor, paying one-time costs up front.\n     *\n     * The first inference of a session is not representative: WebGPU compiles its\n     * shaders on it and the WASM backend faults in its arenas. Calling this while\n     * a loading spinner is still up moves that cost somewhere the user is already\n     * waiting — which matters most for a classifier running as the last step of\n     * an on-device analysis, where the delay lands right before the answer shows.\n     *\n     * @param runs How many warm-up inferences to run. One is enough for WASM;\n     *   WebGPU sometimes settles on the second.\n     */\n    async warmup(runs: number = 1): Promise<void> {\n        const [tw, th] = this._inputSize;\n        for (let i = 0; i < runs; i++) {\n            const tensor = toFloat32Tensor(zeroTensorData(tw, th), [1, 3, th, tw]);\n            await this._session.run({ [this._session.inputName]: tensor });\n        }\n    }\n\n    /**\n     * The fused preprocessing pipeline, built on first use.\n     *\n     * Lazily, because constructing it reserves the output buffer: a task built in\n     * an environment without a canvas implementation stays constructible, and only\n     * fails if it is actually asked to preprocess something.\n     */\n    private get _pipeline(): ResizePipeline {\n        if (this._pipelineCache === null) {\n            const [tw, th] = this._inputSize;\n            this._pipelineCache = new ResizePipeline(tw, th, this._mean, this._std);\n        }\n        return this._pipelineCache;\n    }\n\n    /**\n     * Load the model, resolve labels, and settle the preprocessing.\n     *\n     * @param model The model source — a URL, or the bytes.\n     * @param options Labels, input size, normalization, and session options.\n     * @throws {RangeError} If `normalization` names an unknown preset, or names\n     *   one while `mean`/`std` are also given.\n     */\n    static async create(model: ModelSource, options: ClassifierOptions = {}): Promise<Classifier> {\n        const session = await OrtSession.create(model, options);\n        const normalization = resolveNormalization(session.metadata, {\n            normalization: options.normalization,\n            mean: options.mean,\n            std: options.std,\n        });\n        const numClasses =\n            options.numClasses ?? classificationNumClasses(session.outputShape) ?? undefined;\n        const labels = resolveLabels(options.labels ?? modelNames(session.metadata), {\n            numClasses,\n        });\n        const names: Record<number, string> = {};\n        for (let i = 0; i < labels.length; i++) {\n            names[i] = labels[i] as string;\n        }\n        return new Classifier(\n            session,\n            labels,\n            names,\n            resolveInputSize({\n                graphShape: session.inputShape,\n                requested: options.inputSize,\n                fallback: [224, 224],\n            }),\n            normalization.mean,\n            normalization.std,\n            options.applySoftmax ?? !isUltralyticsClassifier(session.metadata),\n            normalization.name,\n        );\n    }\n\n    /** Class labels indexed by class id. */\n    get labels(): readonly string[] {\n        return this._labels;\n    }\n\n    /** Class id → class name dict (matches Ultralytics' `model.names`). */\n    get names(): Readonly<Record<number, string>> {\n        return this._names;\n    }\n\n    /**\n     * The `[width, height]` this task preprocesses to.\n     *\n     * Resolved at creation time from the model's graph when it declares a static\n     * input, so reading it back tells you the resolution inference really runs at\n     * — not merely what was requested.\n     */\n    get inputSize(): readonly [number, number] {\n        return this._inputSize;\n    }\n\n    /** Number of classes the model can predict. */\n    get numClasses(): number {\n        return this._labels.length;\n    }\n\n    /** Alias for {@link predict} (parity with PyTorch `nn.Module.__call__`). */\n    async call(\n        image: ImageInput,\n        options: ClassifierPredictOptions = {},\n    ): Promise<ClassificationResults[]> {\n        return this.predict(image, options);\n    }\n\n    /** Run classification on a single image. */\n    async predict(\n        image: ImageInput,\n        options: ClassifierPredictOptions = {},\n    ): Promise<ClassificationResults[]> {\n        const timer = new SpeedTimer();\n        const path = typeof image === \"string\" ? image : null;\n        const original = await loadImage(image);\n        timer.stage(\"load\");\n        const tensor = this._preprocess(original);\n        timer.stage(\"preprocess\");\n        const outputs = await this._session.run({ [this._session.inputName]: tensor });\n        this._pipeline.release();\n        timer.stage(\"inference\");\n        const firstOutputName = this._session.outputNames[0];\n        if (firstOutputName === undefined) {\n            throw new Error(\"Classifier model has no outputs.\");\n        }\n        const raw = outputs[firstOutputName];\n        if (raw === undefined) {\n            throw new Error(\n                `Classifier model output ${firstOutputName} missing from run() result.`,\n            );\n        }\n        const fullProbs = this._postprocess(raw.data as Float32Array);\n\n        const { indices, values } = topK(fullProbs, options.topK ?? null);\n        const probabilities: ClassProbability[] = [];\n        for (let i = 0; i < indices.length; i++) {\n            const id = indices[i] as number;\n            const className = this._labels[id] ?? `class_${id}`;\n            probabilities.push({\n                classId: id,\n                className,\n                probability: values[i] as number,\n                cls: id,\n                name: className,\n                conf: values[i] as number,\n            });\n        }\n        if (probabilities.length === 0) {\n            throw new Error(\"Classifier produced no probabilities (empty output).\");\n        }\n\n        const top = probabilities[0] as ClassProbability;\n        const result: ClassificationResult = {\n            classId: top.classId,\n            className: top.className,\n            confidence: top.probability,\n            cls: top.classId,\n            name: top.className,\n            conf: top.probability,\n            image: original,\n            probabilities,\n        };\n\n        const orig: readonly [number, number] = [original.height, original.width];\n        const probs = new Probs(fullProbs);\n        timer.stage(\"postprocess\");\n        return [\n            new ClassificationResults(\n                probs,\n                result,\n                this._names,\n                original,\n                orig,\n                path,\n                timer.speed(),\n            ),\n        ];\n    }\n\n    private _preprocess(image: RGBImage): ort.Tensor {\n        const [tw, th] = this._inputSize;\n        const { data } = this._pipeline.run(image);\n        return toFloat32Tensor(data, [1, 3, th, tw]);\n    }\n\n    private _postprocess(raw: Float32Array): Float32Array {\n        return this._applySoftmax ? softmax(raw) : new Float32Array(raw);\n    }\n}\n"],"mappings":";;;;;;;;;;;;;AAyGA,IAAa,IAAb,MAAa,UAAmB,EAAW;CAGlB;CACA;CACA;CACA;CACA;CACA;CACA;CARrB,YACI,GACA,GACA,GACA,GACA,GACA,GACA,GACA,GACF;EADmB,AAEjB,MAAM,CAAO,GARI,KAAA,UAAA,GACA,KAAA,SAAA,GACA,KAAA,aAAA,GACA,KAAA,QAAA,GACA,KAAA,OAAA,GACA,KAAA,gBAAA,GACA,KAAA,iBAAA;CAGrB;CAUA,IAAI,iBAA0B;EAC1B,OAAO,KAAK;CAChB;CAWA,IAAI,gBAAwB;EACxB,OAAO,KAAK;CAChB;CAEA,iBAAgD;CAchD,MAAM,OAAO,IAAe,GAAkB;EAC1C,IAAM,CAAC,GAAI,KAAM,KAAK;EACtB,KAAK,IAAI,IAAI,GAAG,IAAI,GAAM,KAAK;GAC3B,IAAM,IAAS,EAAgB,EAAe,GAAI,CAAE,GAAG;IAAC;IAAG;IAAG;IAAI;GAAE,CAAC;GACrE,MAAM,KAAK,SAAS,IAAI,GAAG,KAAK,SAAS,YAAY,EAAO,CAAC;EACjE;CACJ;CASA,IAAY,YAA4B;EACpC,IAAI,KAAK,mBAAmB,MAAM;GAC9B,IAAM,CAAC,GAAI,KAAM,KAAK;GACtB,KAAK,iBAAiB,IAAI,EAAe,GAAI,GAAI,KAAK,OAAO,KAAK,IAAI;EAC1E;EACA,OAAO,KAAK;CAChB;CAUA,aAAa,OAAO,GAAoB,IAA6B,CAAC,GAAwB;EAC1F,IAAM,IAAU,MAAM,EAAW,OAAO,GAAO,CAAO,GAChD,IAAgB,EAAqB,EAAQ,UAAU;GACzD,eAAe,EAAQ;GACvB,MAAM,EAAQ;GACd,KAAK,EAAQ;EACjB,CAAC,GACK,IACF,EAAQ,cAAc,EAAyB,EAAQ,WAAW,KAAK,KAAA,GACrE,IAAS,EAAc,EAAQ,UAAU,EAAW,EAAQ,QAAQ,GAAG,EACzE,cACJ,CAAC,GACK,IAAgC,CAAC;EACvC,KAAK,IAAI,IAAI,GAAG,IAAI,EAAO,QAAQ,KAC/B,EAAM,KAAK,EAAO;EAEtB,OAAO,IAAI,EACP,GACA,GACA,GACA,EAAiB;GACb,YAAY,EAAQ;GACpB,WAAW,EAAQ;GACnB,UAAU,CAAC,KAAK,GAAG;EACvB,CAAC,GACD,EAAc,MACd,EAAc,KACd,EAAQ,gBAAgB,CAAC,EAAwB,EAAQ,QAAQ,GACjE,EAAc,IAClB;CACJ;CAGA,IAAI,SAA4B;EAC5B,OAAO,KAAK;CAChB;CAGA,IAAI,QAA0C;EAC1C,OAAO,KAAK;CAChB;CASA,IAAI,YAAuC;EACvC,OAAO,KAAK;CAChB;CAGA,IAAI,aAAqB;EACrB,OAAO,KAAK,QAAQ;CACxB;CAGA,MAAM,KACF,GACA,IAAoC,CAAC,GACL;EAChC,OAAO,KAAK,QAAQ,GAAO,CAAO;CACtC;CAGA,MAAM,QACF,GACA,IAAoC,CAAC,GACL;EAChC,IAAM,IAAQ,IAAI,EAAW,GACvB,IAAO,OAAO,KAAU,WAAW,IAAQ,MAC3C,IAAW,MAAM,EAAU,CAAK;EACtC,EAAM,MAAM,MAAM;EAClB,IAAM,IAAS,KAAK,YAAY,CAAQ;EACxC,EAAM,MAAM,YAAY;EACxB,IAAM,IAAU,MAAM,KAAK,SAAS,IAAI,GAAG,KAAK,SAAS,YAAY,EAAO,CAAC;EAE7E,AADA,KAAK,UAAU,QAAQ,GACvB,EAAM,MAAM,WAAW;EACvB,IAAM,IAAkB,KAAK,SAAS,YAAY;EAClD,IAAI,MAAoB,KAAA,GACpB,MAAU,MAAM,kCAAkC;EAEtD,IAAM,IAAM,EAAQ;EACpB,IAAI,MAAQ,KAAA,GACR,MAAU,MACN,2BAA2B,EAAgB,4BAC/C;EAEJ,IAAM,IAAY,KAAK,aAAa,EAAI,IAAoB,GAEtD,EAAE,YAAS,cAAW,EAAK,GAAW,EAAQ,QAAQ,IAAI,GAC1D,IAAoC,CAAC;EAC3C,KAAK,IAAI,IAAI,GAAG,IAAI,EAAQ,QAAQ,KAAK;GACrC,IAAM,IAAK,EAAQ,IACb,IAAY,KAAK,QAAQ,MAAO,SAAS;GAC/C,EAAc,KAAK;IACf,SAAS;IACT;IACA,aAAa,EAAO;IACpB,KAAK;IACL,MAAM;IACN,MAAM,EAAO;GACjB,CAAC;EACL;EACA,IAAI,EAAc,WAAW,GACzB,MAAU,MAAM,sDAAsD;EAG1E,IAAM,IAAM,EAAc,IACpB,IAA+B;GACjC,SAAS,EAAI;GACb,WAAW,EAAI;GACf,YAAY,EAAI;GAChB,KAAK,EAAI;GACT,MAAM,EAAI;GACV,MAAM,EAAI;GACV,OAAO;GACP;EACJ,GAEM,IAAkC,CAAC,EAAS,QAAQ,EAAS,KAAK,GAClE,IAAQ,IAAI,EAAM,CAAS;EAEjC,OADA,EAAM,MAAM,aAAa,GAClB,CACH,IAAI,EACA,GACA,GACA,KAAK,QACL,GACA,GACA,GACA,EAAM,MAAM,CAChB,CACJ;CACJ;CAEA,YAAoB,GAA6B;EAC7C,IAAM,CAAC,GAAI,KAAM,KAAK,YAChB,EAAE,YAAS,KAAK,UAAU,IAAI,CAAK;EACzC,OAAO,EAAgB,GAAM;GAAC;GAAG;GAAG;GAAI;EAAE,CAAC;CAC/C;CAEA,aAAqB,GAAiC;EAClD,OAAO,KAAK,gBAAgB,EAAQ,CAAG,IAAI,IAAI,aAAa,CAAG;CACnE;AACJ"}
+{"version":3,"file":"classifier.js","names":[],"sources":["../../../src/vision/tasks/classifier.ts"],"sourcesContent":["/** @generated Vendored from @mauriciobenjamin700/ort-vision-sdk-web. Do not hand-edit — regenerate with `npm run vendor:vision`. */\n/**\n * Image classification task using ONNX Runtime Web.\n */\n\nimport type * as ort from \"onnxruntime-web\";\n\nimport { type ModelSource, type OrtSessionOptions, OrtSession } from \"../core/session\";\nimport { SpeedTimer } from \"../core/timing\";\nimport { type ImageInput, loadImageSource } from \"../io/image\";\nimport { classificationNumClasses, resolveInputSize } from \"../core/graph\";\nimport { modelNames } from \"../core/metadata\";\nimport { type LabelSpec, resolveLabels } from \"../labels\";\nimport {\n    type Normalization,\n    isUltralyticsClassifier,\n    resolveNormalization,\n} from \"../normalization\";\nimport { softmax, topK } from \"../postprocess/classification\";\nimport { asFloat32Array } from \"../core/dtypes\";\nimport { toFloat32Tensor } from \"../preprocess/image\";\nimport { ResizePipeline, zeroTensorData } from \"../preprocess/pipeline\";\nimport { ClassificationResults, Probs } from \"../results\";\nimport { VisionTask } from \"./base\";\nimport { type ClassProbability, type ClassificationResult, type RGBImage } from \"../types\";\n\nexport interface ClassifierOptions extends OrtSessionOptions {\n    /**\n     * Class label spec — see {@link resolveLabels}.\n     *\n     * Optional: when omitted, the names the export baked into the model are used\n     * (Ultralytics writes them as `names` in the metadata map). Only when the\n     * model carries none does this fall back to generated `class_<id>` labels.\n     * Passing a spec always wins, for a model whose names are wrong or absent.\n     */\n    readonly labels?: LabelSpec;\n    /**\n     * Number of classes the model can predict.\n     *\n     * Optional: inferred from the classification head's declared output shape\n     * `(B, nc)`. Pass it to validate that the supplied labels match the model.\n     */\n    readonly numClasses?: number;\n    /**\n     * Model input `[width, height]` in pixels.\n     *\n     * Only used when the model's graph leaves its spatial axes dynamic: a graph\n     * that declares a static size always wins, since that is the only shape ONNX\n     * Runtime will accept. Defaults to `[224, 224]`.\n     */\n    readonly inputSize?: readonly [number, number];\n    /**\n     * Which preprocessing this model expects — see {@link Normalization}.\n     *\n     * Defaults to `\"auto\"`, which reads the model's own export metadata and picks\n     * `\"ultralytics\"` (raw `[0, 1]`) for an Ultralytics classification head and\n     * `\"imagenet\"` for everything else.\n     */\n    readonly normalization?: Normalization;\n    /** Per-channel RGB mean, overriding the preset. Defaults to the preset's. */\n    readonly mean?: readonly [number, number, number];\n    /** Per-channel RGB standard deviation, overriding the preset. */\n    readonly std?: readonly [number, number, number];\n    /**\n     * Whether the model's output still needs a softmax.\n     *\n     * Left undefined (the default), this reads the model's metadata and answers\n     * `false` for an Ultralytics classification export, whose graph already ends\n     * in one — applying a second softmax to a probability vector keeps the\n     * ranking but flattens the confidences, so the top-1 stays right while every\n     * number attached to it is wrong. Detection covers that family; for any other\n     * model that already emits probabilities, pass `false` explicitly.\n     */\n    readonly applySoftmax?: boolean;\n}\n\nexport interface ClassifierPredictOptions {\n    /**\n     * If set, the per-class probability list in `results[0].result.probabilities`\n     * is truncated to the top-K entries. The bulk `probs` view always exposes\n     * the full vector.\n     */\n    readonly topK?: number;\n}\n\n/**\n * Image classifier wrapping an ONNX model with ImageNet-style preprocessing.\n *\n * `predict()` returns `Promise<ClassificationResults[]>` (length 1 for a\n * single image), mirroring Ultralytics' API. The envelope exposes a `probs`\n * collection (`top1`, `top1conf`, `top5`, `top5conf`, `data`) plus the\n * legacy per-class probability list with names already resolved.\n *\n * Defaults: 224×224 RGB input, `float32` normalized with ImageNet mean/std,\n * NCHW layout, batch size 1, softmax applied to the raw output.\n *\n * @example\n * ```typescript\n * const clf = await Classifier.create(\"/models/resnet50.onnx\", {\n *   labels: [\"tench\", \"goldfish\", ...]  // 1000 ImageNet labels\n * });\n * const r = (await clf.predict(\"/images/dog.jpg\"))[0];\n * console.log(r.cls, r.conf, r.name);\n * console.log(r.probs.top5, r.probs.top5conf);\n * ```\n */\nexport class Classifier extends VisionTask {\n    private constructor(\n        session: OrtSession,\n        private readonly _labels: readonly string[],\n        private readonly _names: Readonly<Record<number, string>>,\n        private readonly _inputSize: readonly [number, number],\n        private readonly _mean: readonly [number, number, number],\n        private readonly _std: readonly [number, number, number],\n        private readonly _applySoftmax: boolean,\n        private readonly _normalization: string,\n    ) {\n        super(session);\n    }\n\n    /**\n     * Whether a softmax is applied to the model's output before ranking.\n     *\n     * Resolved once at construction. Worth reading when confidences look\n     * compressed: a second softmax over an already-normalized vector leaves the\n     * ordering intact and the numbers meaningless, which is invisible to any check\n     * that only looks at the predicted class.\n     */\n    get appliesSoftmax(): boolean {\n        return this._applySoftmax;\n    }\n\n    /**\n     * Which preprocessing this classifier applies to every image.\n     *\n     * One of the {@link Normalization} preset names, or `\"custom\"` when the caller\n     * supplied `mean`/`std` directly. Worth reading when a model underperforms:\n     * feeding a classifier a differently prepared tensor than it was trained on\n     * degrades it without throwing anything, so \"what does this assume\" is the\n     * first question.\n     */\n    get normalization(): string {\n        return this._normalization;\n    }\n\n    private _pipelineCache: ResizePipeline | null = null;\n\n    /**\n     * Run the model once on a zero tensor, paying one-time costs up front.\n     *\n     * The first inference of a session is not representative: WebGPU compiles its\n     * shaders on it and the WASM backend faults in its arenas. Calling this while\n     * a loading spinner is still up moves that cost somewhere the user is already\n     * waiting — which matters most for a classifier running as the last step of\n     * an on-device analysis, where the delay lands right before the answer shows.\n     *\n     * @param runs How many warm-up inferences to run. One is enough for WASM;\n     *   WebGPU sometimes settles on the second.\n     */\n    async warmup(runs: number = 1): Promise<void> {\n        const [tw, th] = this._inputSize;\n        for (let i = 0; i < runs; i++) {\n            const tensor = toFloat32Tensor(zeroTensorData(tw, th), [1, 3, th, tw]);\n            await this._session.run({ [this._session.inputName]: tensor });\n        }\n    }\n\n    /**\n     * The fused preprocessing pipeline, built on first use.\n     *\n     * Lazily, because constructing it reserves the output buffer: a task built in\n     * an environment without a canvas implementation stays constructible, and only\n     * fails if it is actually asked to preprocess something.\n     */\n    private get _pipeline(): ResizePipeline {\n        if (this._pipelineCache === null) {\n            const [tw, th] = this._inputSize;\n            this._pipelineCache = new ResizePipeline(tw, th, this._mean, this._std);\n        }\n        return this._pipelineCache;\n    }\n\n    /**\n     * Load the model, resolve labels, and settle the preprocessing.\n     *\n     * @param model The model source — a URL, or the bytes.\n     * @param options Labels, input size, normalization, and session options.\n     * @throws {RangeError} If `normalization` names an unknown preset, or names\n     *   one while `mean`/`std` are also given.\n     */\n    static async create(model: ModelSource, options: ClassifierOptions = {}): Promise<Classifier> {\n        const session = await OrtSession.create(model, options);\n        const normalization = resolveNormalization(session.metadata, {\n            normalization: options.normalization,\n            mean: options.mean,\n            std: options.std,\n        });\n        const numClasses =\n            options.numClasses ?? classificationNumClasses(session.outputShape) ?? undefined;\n        const labels = resolveLabels(options.labels ?? modelNames(session.metadata), {\n            numClasses,\n        });\n        const names: Record<number, string> = {};\n        for (let i = 0; i < labels.length; i++) {\n            names[i] = labels[i] as string;\n        }\n        return new Classifier(\n            session,\n            labels,\n            names,\n            resolveInputSize({\n                graphShape: session.inputShape,\n                requested: options.inputSize,\n                fallback: [224, 224],\n            }),\n            normalization.mean,\n            normalization.std,\n            options.applySoftmax ?? !isUltralyticsClassifier(session.metadata),\n            normalization.name,\n        );\n    }\n\n    /** Class labels indexed by class id. */\n    get labels(): readonly string[] {\n        return this._labels;\n    }\n\n    /** Class id → class name dict (matches Ultralytics' `model.names`). */\n    get names(): Readonly<Record<number, string>> {\n        return this._names;\n    }\n\n    /**\n     * The `[width, height]` this task preprocesses to.\n     *\n     * Resolved at creation time from the model's graph when it declares a static\n     * input, so reading it back tells you the resolution inference really runs at\n     * — not merely what was requested.\n     */\n    get inputSize(): readonly [number, number] {\n        return this._inputSize;\n    }\n\n    /** Number of classes the model can predict. */\n    get numClasses(): number {\n        return this._labels.length;\n    }\n\n    /** Alias for {@link predict} (parity with PyTorch `nn.Module.__call__`). */\n    async call(\n        image: ImageInput,\n        options: ClassifierPredictOptions = {},\n    ): Promise<ClassificationResults[]> {\n        return this.predict(image, options);\n    }\n\n    /** Run classification on a single image. */\n    async predict(\n        image: ImageInput,\n        options: ClassifierPredictOptions = {},\n    ): Promise<ClassificationResults[]> {\n        const timer = new SpeedTimer();\n        const path = typeof image === \"string\" ? image : null;\n        const { image: original, canvas } = await loadImageSource(image);\n        timer.stage(\"load\");\n        const { tensor, reused } = this._preprocess(original, canvas);\n        timer.stage(\"preprocess\");\n        let outputs: Record<string, ort.Tensor>;\n        try {\n            outputs = await this._session.run({ [this._session.inputName]: tensor });\n        } finally {\n            this._pipeline.release({ reused });\n        }\n        timer.stage(\"inference\");\n        const firstOutputName = this._session.outputNames[0];\n        if (firstOutputName === undefined) {\n            throw new Error(\"Classifier model has no outputs.\");\n        }\n        const raw = outputs[firstOutputName];\n        if (raw === undefined) {\n            throw new Error(\n                `Classifier model output ${firstOutputName} missing from run() result.`,\n            );\n        }\n        const fullProbs = this._postprocess(asFloat32Array(raw.data));\n\n        const { indices, values } = topK(fullProbs, options.topK ?? null);\n        const probabilities: ClassProbability[] = [];\n        for (let i = 0; i < indices.length; i++) {\n            const id = indices[i] as number;\n            const className = this._labels[id] ?? `class_${id}`;\n            probabilities.push({\n                classId: id,\n                className,\n                probability: values[i] as number,\n                cls: id,\n                name: className,\n                conf: values[i] as number,\n            });\n        }\n        if (probabilities.length === 0) {\n            throw new Error(\"Classifier produced no probabilities (empty output).\");\n        }\n\n        const top = probabilities[0] as ClassProbability;\n        const result: ClassificationResult = {\n            classId: top.classId,\n            className: top.className,\n            confidence: top.probability,\n            cls: top.classId,\n            name: top.className,\n            conf: top.probability,\n            image: original,\n            probabilities,\n        };\n\n        const orig: readonly [number, number] = [original.height, original.width];\n        const probs = new Probs(fullProbs);\n        timer.stage(\"postprocess\");\n        return [\n            new ClassificationResults(\n                probs,\n                result,\n                this._names,\n                original,\n                orig,\n                path,\n                timer.speed(),\n            ),\n        ];\n    }\n\n    /**\n     * Resize and pack the image into the tensor the model expects.\n     *\n     * @param image The decoded input.\n     * @param canvas The opaque canvas it was decoded on, when there is one — see\n     *   {@link ResizePipeline.run}.\n     */\n    private _preprocess(\n        image: RGBImage,\n        canvas: CanvasImageSource | null,\n    ): { tensor: ort.Tensor; reused: boolean } {\n        const [tw, th] = this._inputSize;\n        const { data, reused } = this._pipeline.run(image, canvas);\n        return { tensor: toFloat32Tensor(data, [1, 3, th, tw]), reused };\n    }\n\n    private _postprocess(raw: Float32Array): Float32Array {\n        return this._applySoftmax ? softmax(raw) : new Float32Array(raw);\n    }\n}\n"],"mappings":";;;;;;;;;;;;;;AA0GA,IAAa,IAAb,MAAa,UAAmB,EAAW;CAGlB;CACA;CACA;CACA;CACA;CACA;CACA;CARrB,YACI,GACA,GACA,GACA,GACA,GACA,GACA,GACA,GACF;EADmB,AAEjB,MAAM,CAAO,GARI,KAAA,UAAA,GACA,KAAA,SAAA,GACA,KAAA,aAAA,GACA,KAAA,QAAA,GACA,KAAA,OAAA,GACA,KAAA,gBAAA,GACA,KAAA,iBAAA;CAGrB;CAUA,IAAI,iBAA0B;EAC1B,OAAO,KAAK;CAChB;CAWA,IAAI,gBAAwB;EACxB,OAAO,KAAK;CAChB;CAEA,iBAAgD;CAchD,MAAM,OAAO,IAAe,GAAkB;EAC1C,IAAM,CAAC,GAAI,KAAM,KAAK;EACtB,KAAK,IAAI,IAAI,GAAG,IAAI,GAAM,KAAK;GAC3B,IAAM,IAAS,EAAgB,EAAe,GAAI,CAAE,GAAG;IAAC;IAAG;IAAG;IAAI;GAAE,CAAC;GACrE,MAAM,KAAK,SAAS,IAAI,GAAG,KAAK,SAAS,YAAY,EAAO,CAAC;EACjE;CACJ;CASA,IAAY,YAA4B;EACpC,IAAI,KAAK,mBAAmB,MAAM;GAC9B,IAAM,CAAC,GAAI,KAAM,KAAK;GACtB,KAAK,iBAAiB,IAAI,EAAe,GAAI,GAAI,KAAK,OAAO,KAAK,IAAI;EAC1E;EACA,OAAO,KAAK;CAChB;CAUA,aAAa,OAAO,GAAoB,IAA6B,CAAC,GAAwB;EAC1F,IAAM,IAAU,MAAM,EAAW,OAAO,GAAO,CAAO,GAChD,IAAgB,EAAqB,EAAQ,UAAU;GACzD,eAAe,EAAQ;GACvB,MAAM,EAAQ;GACd,KAAK,EAAQ;EACjB,CAAC,GACK,IACF,EAAQ,cAAc,EAAyB,EAAQ,WAAW,KAAK,KAAA,GACrE,IAAS,EAAc,EAAQ,UAAU,EAAW,EAAQ,QAAQ,GAAG,EACzE,cACJ,CAAC,GACK,IAAgC,CAAC;EACvC,KAAK,IAAI,IAAI,GAAG,IAAI,EAAO,QAAQ,KAC/B,EAAM,KAAK,EAAO;EAEtB,OAAO,IAAI,EACP,GACA,GACA,GACA,EAAiB;GACb,YAAY,EAAQ;GACpB,WAAW,EAAQ;GACnB,UAAU,CAAC,KAAK,GAAG;EACvB,CAAC,GACD,EAAc,MACd,EAAc,KACd,EAAQ,gBAAgB,CAAC,EAAwB,EAAQ,QAAQ,GACjE,EAAc,IAClB;CACJ;CAGA,IAAI,SAA4B;EAC5B,OAAO,KAAK;CAChB;CAGA,IAAI,QAA0C;EAC1C,OAAO,KAAK;CAChB;CASA,IAAI,YAAuC;EACvC,OAAO,KAAK;CAChB;CAGA,IAAI,aAAqB;EACrB,OAAO,KAAK,QAAQ;CACxB;CAGA,MAAM,KACF,GACA,IAAoC,CAAC,GACL;EAChC,OAAO,KAAK,QAAQ,GAAO,CAAO;CACtC;CAGA,MAAM,QACF,GACA,IAAoC,CAAC,GACL;EAChC,IAAM,IAAQ,IAAI,EAAW,GACvB,IAAO,OAAO,KAAU,WAAW,IAAQ,MAC3C,EAAE,OAAO,GAAU,cAAW,MAAM,EAAgB,CAAK;EAC/D,EAAM,MAAM,MAAM;EAClB,IAAM,EAAE,WAAQ,cAAW,KAAK,YAAY,GAAU,CAAM;EAC5D,EAAM,MAAM,YAAY;EACxB,IAAI;EACJ,IAAI;GACA,IAAU,MAAM,KAAK,SAAS,IAAI,GAAG,KAAK,SAAS,YAAY,EAAO,CAAC;EAC3E,UAAU;GACN,KAAK,UAAU,QAAQ,EAAE,UAAO,CAAC;EACrC;EACA,EAAM,MAAM,WAAW;EACvB,IAAM,IAAkB,KAAK,SAAS,YAAY;EAClD,IAAI,MAAoB,KAAA,GACpB,MAAU,MAAM,kCAAkC;EAEtD,IAAM,IAAM,EAAQ;EACpB,IAAI,MAAQ,KAAA,GACR,MAAU,MACN,2BAA2B,EAAgB,4BAC/C;EAEJ,IAAM,IAAY,KAAK,aAAa,EAAe,EAAI,IAAI,CAAC,GAEtD,EAAE,YAAS,cAAW,EAAK,GAAW,EAAQ,QAAQ,IAAI,GAC1D,IAAoC,CAAC;EAC3C,KAAK,IAAI,IAAI,GAAG,IAAI,EAAQ,QAAQ,KAAK;GACrC,IAAM,IAAK,EAAQ,IACb,IAAY,KAAK,QAAQ,MAAO,SAAS;GAC/C,EAAc,KAAK;IACf,SAAS;IACT;IACA,aAAa,EAAO;IACpB,KAAK;IACL,MAAM;IACN,MAAM,EAAO;GACjB,CAAC;EACL;EACA,IAAI,EAAc,WAAW,GACzB,MAAU,MAAM,sDAAsD;EAG1E,IAAM,IAAM,EAAc,IACpB,IAA+B;GACjC,SAAS,EAAI;GACb,WAAW,EAAI;GACf,YAAY,EAAI;GAChB,KAAK,EAAI;GACT,MAAM,EAAI;GACV,MAAM,EAAI;GACV,OAAO;GACP;EACJ,GAEM,IAAkC,CAAC,EAAS,QAAQ,EAAS,KAAK,GAClE,IAAQ,IAAI,EAAM,CAAS;EAEjC,OADA,EAAM,MAAM,aAAa,GAClB,CACH,IAAI,EACA,GACA,GACA,KAAK,QACL,GACA,GACA,GACA,EAAM,MAAM,CAChB,CACJ;CACJ;CASA,YACI,GACA,GACuC;EACvC,IAAM,CAAC,GAAI,KAAM,KAAK,YAChB,EAAE,SAAM,cAAW,KAAK,UAAU,IAAI,GAAO,CAAM;EACzD,OAAO;GAAE,QAAQ,EAAgB,GAAM;IAAC;IAAG;IAAG;IAAI;GAAE,CAAC;GAAG;EAAO;CACnE;CAEA,aAAqB,GAAiC;EAClD,OAAO,KAAK,gBAAgB,EAAQ,CAAG,IAAI,IAAI,aAAa,CAAG;CACnE;AACJ"}

package/dist/vision/tasks/detectClassify.cjs

@@ -1,2 +1,2 @@
-const e=require("../core/exceptions.cjs"),t=require("../types.cjs"),n=require("../core/timing.cjs"),r=require("../results.cjs"),i=require("../labels.cjs"),a=require("../core/session.cjs"),o=require("../io/image.cjs"),s=require("../fusion.cjs"),c=require("../preprocess/image.cjs"),l=require("../preprocess/pipeline.cjs"),u=require("../postprocess/classification.cjs"),d=require("./base.cjs");var f=class f extends d.VisionTask{_spec;_labels;_names;_classifierLabels;_classifierNames;_raiseOnEmpty;constructor(e,t,n,r,i,a,o){super(e),this._spec=t,this._labels=n,this._names=r,this._classifierLabels=i,this._classifierNames=a,this._raiseOnEmpty=o}_pipelineCache=null;async warmup(e=1){let[t,n]=this._spec.inputSize;for(let r=0;r<e;r++){let e={[s.INPUT_IMAGE]:c.toFloat32Tensor(l.zeroTensorData(t,n),[1,3,n,t])};this._spec.needsSourceImage&&(e[s.INPUT_SOURCE]=c.toFloat32Tensor(l.zeroTensorData(t,n),[1,3,n,t]),e[s.INPUT_SCALE]=c.toFloat32Tensor(new Float32Array([1]),[1]),e[s.INPUT_PAD]=c.toFloat32Tensor(new Float32Array([0,0]),[2])),await this._session.run(e)}}get _pipeline(){if(this._pipelineCache===null){let[e,t]=this._spec.inputSize;this._pipelineCache=new l.LetterboxPipeline(e,t)}return this._pipelineCache}static async create(t,n={}){let r=await a.OrtSession.create(t,n),o=s.readFusionSpec(r.metadata);if(o===null)throw new e.FusionError(`This model carries no fused-pipeline metadata, so DetectClassify cannot tell how to drive it. Build one with ort_vision_sdk.compose.fuse_detect_classify, or load a plain model with Detector/Classifier instead.`);let c=i.resolveLabels(n.labels??o.detectorNames??`coco`),l=i.resolveLabels(n.classifierLabels??o.classifierNames,{numClasses:p(r)??void 0});return new f(r,o,c,m(c),l,m(l),n.raiseOnEmpty??!1)}get spec(){return this._spec}get inputSize(){return this._spec.inputSize}get labels(){return this._labels}get names(){return this._names}get classifierLabels(){return this._classifierLabels}get classifierNames(){return this._classifierNames}async call(e,t={}){return this.predict(e,t)}async predict(e,t={}){let i=new n.SpeedTimer,a=typeof e==`string`?e:null,c=await o.loadImage(e);i.stage(`load`);let{feeds:l,scale:u,padLeft:f,padTop:p}=this._preprocess(c);i.stage(`preprocess`);let m=await this._session.run(l);this._pipeline.release(),i.stage(`inference`);let v=h(m,s.OUTPUT_PROBS),S=g(m,s.OUTPUT_BOXES),C=g(m,s.OUTPUT_SCORES),w=_(m,s.OUTPUT_CLASSES),T=v.data,E=_(m,`num_detections`)[0]??0,D=Math.min(E,Math.floor(S.length/4)),O=v.dims[v.dims.length-1]??0,k=t.classes===void 0?null:new Set(t.classes),A=t.confThreshold??0,j=[];for(let e=0;e<D;e++){let n=w[e]??0,r=C[e]??0;if(r<A||k!==null&&!k.has(n))continue;let i=this._toOriginal(S,e,{scale:u,padLeft:f,padTop:p,original:c}),a=y(c,i);j.push(b(n,this._names[n]??`class_${n}`,r,i,a,this._classify(T.subarray(e*O,(e+1)*O),a,t.topK)))}d.requireDetections(j.length,{raiseOnEmpty:t.raiseOnEmpty??this._raiseOnEmpty,confThreshold:Math.max(A,this._spec.confThreshold),classes:t.classes,path:a});let M=[c.height,c.width];return i.stage(`postprocess`),[new r.DetectClassifyResults(x(j,M),j,this._names,this._classifierNames,c,M,a,i.speed())]}_preprocess(e){let[t,n]=this._spec.inputSize,r=this._pipeline.run(e),i={[s.INPUT_IMAGE]:c.toFloat32Tensor(r.data,[1,3,n,t])};return this._spec.needsSourceImage&&(i[s.INPUT_SOURCE]=v(e),i[s.INPUT_SCALE]=c.toFloat32Tensor(new Float32Array([r.scale]),[1]),i[s.INPUT_PAD]=c.toFloat32Tensor(new Float32Array([r.padLeft,r.padTop]),[2])),{feeds:i,scale:r.scale,padLeft:r.padLeft,padTop:r.padTop}}_toOriginal(e,n,r){let{scale:i,padLeft:a,padTop:o,original:s}=r,c=t=>e[n*4+t]??0,l=e=>Math.min(Math.max(e,0),s.width),u=e=>Math.min(Math.max(e,0),s.height);return new t.BoundingBox(l((c(0)-a)/i),u((c(1)-o)/i),l((c(2)-a)/i),u((c(3)-o)/i))}_classify(e,t,n){let r=this._spec.applySoftmax?u.softmax(e):e,{indices:i,values:a}=u.topK(r,n??null),o=[];for(let e=0;e<i.length;e++){let t=i[e]??0,n=a[e]??0,r=this._classifierLabels[t]??`class_${t}`;o.push({classId:t,className:r,probability:n,cls:t,name:r,conf:n})}let s=o[0]??{classId:0,className:`class_0`,probability:0,cls:0,name:`class_0`,conf:0};return{classId:s.classId,className:s.className,confidence:s.probability,cls:s.classId,name:s.className,conf:s.probability,image:t,probabilities:o}}};function p(e){let t=e.outputNames.indexOf(s.OUTPUT_PROBS);if(t<0)return null;let n=e.outputShapes[t];return n===void 0||n.length===0?null:n[n.length-1]??null}function m(e){let t={};for(let n=0;n<e.length;n++)t[n]=e[n];return t}function h(t,n){let r=t[n];if(r===void 0)throw new e.FusionError(`The fused pipeline is missing its '${n}' output.`);return r}function g(e,t){return h(e,t).data}function _(e,t){let n=h(e,t).data,r=[];for(let e=0;e<n.length;e++)r.push(Number(n[e]));return r}function v(e){let t=c.toCHW(c.toFloat32(e),e.width,e.height,3);return c.toFloat32Tensor(t,[1,3,e.height,e.width])}function y(e,n){let[r,i,a,o]=n.asIntXyxy(),s=Math.max(0,r),c=Math.max(0,i),l=Math.min(e.width,a),u=Math.min(e.height,o);if(l<=s||u<=c)return new t.RGBImage(new Uint8Array,0,0);let d=l-s,f=u-c,p=new Uint8Array(d*f*3);for(let t=0;t<f;t++){let n=((c+t)*e.width+s)*3;p.set(e.data.subarray(n,n+d*3),t*d*3)}return new t.RGBImage(p,d,f)}function b(e,t,n,r,i,a){return{classId:e,className:t,confidence:n,bbox:r,cls:e,name:t,conf:n,box:r,croppedImage:i,classification:a}}function x(e,t){let n=new Float32Array(e.length*4),i=new Int32Array(e.length),a=new Float32Array(e.length);return e.forEach((e,t)=>{let[r,o,s,c]=e.bbox.xyxy;n[t*4]=r,n[t*4+1]=o,n[t*4+2]=s,n[t*4+3]=c,i[t]=e.classId,a[t]=e.confidence}),new r.Boxes(n,i,a,t)}exports.DetectClassify=f;
+const e=require("../core/exceptions.cjs"),t=require("../types.cjs"),n=require("../core/timing.cjs"),r=require("../results.cjs"),i=require("../labels.cjs"),a=require("../core/dtypes.cjs"),o=require("../core/session.cjs"),s=require("../io/image.cjs"),c=require("../fusion.cjs"),l=require("../preprocess/image.cjs"),u=require("../preprocess/pipeline.cjs"),d=require("../postprocess/classification.cjs"),f=require("./base.cjs");var p=class p extends f.VisionTask{_spec;_labels;_names;_classifierLabels;_classifierNames;_raiseOnEmpty;constructor(e,t,n,r,i,a,o){super(e),this._spec=t,this._labels=n,this._names=r,this._classifierLabels=i,this._classifierNames=a,this._raiseOnEmpty=o}_pipelineCache=null;async warmup(e=1){let[t,n]=this._spec.inputSize;for(let r=0;r<e;r++){let e={[c.INPUT_IMAGE]:l.toFloat32Tensor(u.zeroTensorData(t,n),[1,3,n,t])};this._spec.needsSourceImage&&(e[c.INPUT_SOURCE]=l.toFloat32Tensor(u.zeroTensorData(t,n),[1,3,n,t]),e[c.INPUT_SCALE]=l.toFloat32Tensor(new Float32Array([1]),[1]),e[c.INPUT_PAD]=l.toFloat32Tensor(new Float32Array([0,0]),[2])),await this._session.run(e)}}get _pipeline(){if(this._pipelineCache===null){let[e,t]=this._spec.inputSize;this._pipelineCache=new u.LetterboxPipeline(e,t)}return this._pipelineCache}static async create(t,n={}){let r=await o.OrtSession.create(t,n),a=c.readFusionSpec(r.metadata);if(a===null)throw new e.FusionError(`This model carries no fused-pipeline metadata, so DetectClassify cannot tell how to drive it. Build one with ort_vision_sdk.compose.fuse_detect_classify, or load a plain model with Detector/Classifier instead.`);let s=i.resolveLabels(n.labels??a.detectorNames??`coco`),l=i.resolveLabels(n.classifierLabels??a.classifierNames,{numClasses:m(r)??void 0});return new p(r,a,s,h(s),l,h(l),n.raiseOnEmpty??!1)}get spec(){return this._spec}get inputSize(){return this._spec.inputSize}get labels(){return this._labels}get names(){return this._names}get classifierLabels(){return this._classifierLabels}get classifierNames(){return this._classifierNames}async call(e,t={}){return this.predict(e,t)}async predict(e,t={}){let i=new n.SpeedTimer,o=typeof e==`string`?e:null,{image:l,canvas:u}=await s.loadImageSource(e);i.stage(`load`);let{feeds:d,scale:p,padLeft:m,padTop:h,reused:y}=this._preprocess(l,u);i.stage(`preprocess`);let C;try{C=await this._session.run(d)}finally{this._pipeline.release({reused:y})}i.stage(`inference`);let w=g(C,c.OUTPUT_PROBS),T=_(C,c.OUTPUT_BOXES),E=_(C,c.OUTPUT_SCORES),D=v(C,c.OUTPUT_CLASSES),O=a.asFloat32Array(w.data),k=v(C,`num_detections`)[0]??0,A=this._spec.hasMasks?_(C,c.OUTPUT_MASKS):null,[j,M]=this._spec.cropSize,N=Math.min(k,Math.floor(T.length/4)),P=w.dims[w.dims.length-1]??0,F=t.classes===void 0?null:new Set(t.classes),I=t.confThreshold??0,L=[];for(let e=0;e<N;e++){let n=D[e]??0,r=E[e]??0;if(r<I||F!==null&&!F.has(n))continue;let i=this._toOriginal(T,e,{scale:p,padLeft:m,padTop:h,original:l}),a=f.cropToBox(l,i);L.push(b(n,this._names[n]??`class_${n}`,r,i,a,this._classify(O.subarray(e*P,(e+1)*P),a,t.topK),A===null?null:x(A,e,j,M,a)))}f.requireDetections(L.length,{raiseOnEmpty:t.raiseOnEmpty??this._raiseOnEmpty,confThreshold:Math.max(I,this._spec.confThreshold),classes:t.classes,path:o});let R=[l.height,l.width];return i.stage(`postprocess`),[new r.DetectClassifyResults(S(L,R),L,this._names,this._classifierNames,l,R,o,i.speed())]}_preprocess(e,t){let[n,r]=this._spec.inputSize,i=this._pipeline.run(e,t),a={[c.INPUT_IMAGE]:l.toFloat32Tensor(i.data,[1,3,r,n])};return this._spec.needsSourceImage&&(a[c.INPUT_SOURCE]=y(e),a[c.INPUT_SCALE]=l.toFloat32Tensor(new Float32Array([i.scale]),[1]),a[c.INPUT_PAD]=l.toFloat32Tensor(new Float32Array([i.padLeft,i.padTop]),[2])),{feeds:a,scale:i.scale,padLeft:i.padLeft,padTop:i.padTop,reused:i.reused}}_toOriginal(e,n,r){let{scale:i,padLeft:a,padTop:o,original:s}=r,c=t=>e[n*4+t]??0,l=e=>Math.min(Math.max(e,0),s.width),u=e=>Math.min(Math.max(e,0),s.height);return new t.BoundingBox(l((c(0)-a)/i),u((c(1)-o)/i),l((c(2)-a)/i),u((c(3)-o)/i))}_classify(e,t,n){let r=this._spec.applySoftmax?d.softmax(e):e,{indices:i,values:a}=d.topK(r,n??null),o=[];for(let e=0;e<i.length;e++){let t=i[e]??0,n=a[e]??0,r=this._classifierLabels[t]??`class_${t}`;o.push({classId:t,className:r,probability:n,cls:t,name:r,conf:n})}let s=o[0]??{classId:0,className:`class_0`,probability:0,cls:0,name:`class_0`,conf:0};return{classId:s.classId,className:s.className,confidence:s.probability,cls:s.classId,name:s.className,conf:s.probability,image:t,probabilities:o}}};function m(e){let t=e.outputNames.indexOf(c.OUTPUT_PROBS);if(t<0)return null;let n=e.outputShapes[t];return n===void 0||n.length===0?null:n[n.length-1]??null}function h(e){let t={};for(let n=0;n<e.length;n++)t[n]=e[n];return t}function g(t,n){let r=t[n];if(r===void 0)throw new e.FusionError(`The fused pipeline is missing its '${n}' output.`);return r}function _(e,t){return a.asFloat32Array(g(e,t).data)}function v(e,t){let n=g(e,t).data,r=[];for(let e=0;e<n.length;e++)r.push(Number(n[e]));return r}function y(e){let t=l.toCHW(l.toFloat32(e),e.width,e.height,3);return l.toFloat32Tensor(t,[1,3,e.height,e.width])}function b(e,t,n,r,i,a,o){return{classId:e,className:t,confidence:n,bbox:r,cls:e,name:t,conf:n,box:r,croppedImage:i,classification:a,mask:o}}function x(e,n,r,i,a){let{width:o,height:s}=a;if(o<1||s<1)return null;let c=r*i,l=e.subarray(n*c,(n+1)*c),u=new Uint8Array(o*s);for(let e=0;e<s;e++){let t=Math.min(Math.floor(e*i/s),i-1);for(let n=0;n<o;n++){let i=Math.min(Math.floor(n*r/o),r-1);u[e*o+n]=(l[t*r+i]??0)>.5?255:0}}return new t.Mask(u,o,s)}function S(e,t){let n=new Float32Array(e.length*4),i=new Int32Array(e.length),a=new Float32Array(e.length);return e.forEach((e,t)=>{let[r,o,s,c]=e.bbox.xyxy;n[t*4]=r,n[t*4+1]=o,n[t*4+2]=s,n[t*4+3]=c,i[t]=e.classId,a[t]=e.confidence}),new r.Boxes(n,i,a,t)}exports.DetectClassify=p,exports.maskToBox=x;
 //# sourceMappingURL=detectClassify.cjs.map

package/dist/vision/tasks/detectClassify.cjs.map

@@ -1 +1 @@
-{"version":3,"file":"detectClassify.cjs","names":[],"sources":["../../../src/vision/tasks/detectClassify.ts"],"sourcesContent":["/** @generated Vendored from @mauriciobenjamin700/ort-vision-sdk-web. Do not hand-edit — regenerate with `npm run vendor:vision`. */\n/**\n * Run a fused detect→classify pipeline in the browser.\n *\n * The file this loads was built by the Python SDK's `ort_vision_sdk.compose`,\n * and it already contains both models plus the crop-and-resize bridge between\n * them. That matters far more in a tab than on a server: two models mean two\n * `.onnx` downloads, two WASM/WebGPU session initializations, and a per-crop\n * round trip through JavaScript to slice, resize and restack the regions before\n * the second model can see them. A fused pipeline has one download, one session\n * and no round trip — the crops are produced and consumed inside the graph.\n */\n\nimport type * as ort from \"onnxruntime-web\";\n\nimport { FusionError } from \"../core/exceptions\";\nimport { type ModelSource, type OrtSessionOptions, OrtSession } from \"../core/session\";\nimport { SpeedTimer } from \"../core/timing\";\nimport {\n    INPUT_IMAGE,\n    INPUT_PAD,\n    INPUT_SCALE,\n    INPUT_SOURCE,\n    OUTPUT_BOXES,\n    OUTPUT_CLASSES,\n    OUTPUT_NUM_DETECTIONS,\n    OUTPUT_PROBS,\n    OUTPUT_SCORES,\n    type FusionSpec,\n    readFusionSpec,\n} from \"../fusion\";\nimport { type ImageInput, loadImage } from \"../io/image\";\nimport { type LabelSpec, resolveLabels } from \"../labels\";\nimport { softmax, topK } from \"../postprocess/classification\";\nimport { toCHW, toFloat32, toFloat32Tensor } from \"../preprocess/image\";\nimport { LetterboxPipeline, zeroTensorData } from \"../preprocess/pipeline\";\nimport { Boxes, DetectClassifyResults } from \"../results\";\nimport {\n    BoundingBox,\n    RGBImage,\n    type ClassProbability,\n    type ClassificationResult,\n    type DetectionResult,\n} from \"../types\";\nimport { VisionTask, requireDetections } from \"./base\";\n\nexport interface DetectClassifyOptions extends OrtSessionOptions {\n    /**\n     * Class label spec for the **detection** stage — see {@link resolveLabels}.\n     * Defaults to the names recorded at fusion time, falling back to the COCO\n     * 80-class preset when the fusion recorded none.\n     */\n    readonly labels?: LabelSpec;\n    /**\n     * Class label spec for the **classification** stage. Defaults to the recorded\n     * names, falling back to generated `class_<id>` names.\n     */\n    readonly classifierLabels?: LabelSpec;\n    /**\n     * If `true`, a run that finds nothing throws {@link NoDetectionsError}\n     * instead of returning an empty envelope. Default `false`, because looking\n     * and finding nothing is a successful inference. Turn it on when an empty\n     * result means the surrounding pipeline should stop rather than carry on with\n     * zero rows. Can be overridden per `predict` call.\n     */\n    readonly raiseOnEmpty?: boolean;\n}\n\nexport interface DetectClassifyPredictOptions {\n    /**\n     * Drop detections scoring below this. The graph's own NMS threshold was fixed\n     * at fusion time and cannot be lowered here — this only filters further.\n     */\n    readonly confThreshold?: number;\n    /** If set, keep only detections whose detector `classId` is in this list. */\n    readonly classes?: readonly number[];\n    /** Truncate each detection's `classification.probabilities` to its top-k entries. */\n    readonly topK?: number;\n    /** Override the constructor's `raiseOnEmpty` setting for this call. */\n    readonly raiseOnEmpty?: boolean;\n}\n\n/**\n * Detector and classifier running as a single ONNX model.\n *\n * Everything the pipeline needs to know about itself — the resolution to\n * letterbox to, whether it wants the full-resolution image as well, whether its\n * classifier output still needs a softmax, the class names of both stages — was\n * written into the file at fusion time and is read back here. Nothing is\n * restated on the JavaScript side, so nothing can drift out of step with the\n * Python side that built it.\n *\n * @example\n * ```typescript\n * const pipeline = await DetectClassify.create(\"/models/pipeline.onnx\");\n * const result = (await pipeline.predict(\"/images/flock.jpg\"))[0];\n * for (const detection of result) {\n *   console.log(detection.name, detection.conf, detection.classification?.name);\n * }\n * ```\n */\nexport class DetectClassify extends VisionTask {\n    private constructor(\n        session: OrtSession,\n        private readonly _spec: FusionSpec,\n        private readonly _labels: readonly string[],\n        private readonly _names: Readonly<Record<number, string>>,\n        private readonly _classifierLabels: readonly string[],\n        private readonly _classifierNames: Readonly<Record<number, string>>,\n        private readonly _raiseOnEmpty: boolean,\n    ) {\n        super(session);\n    }\n\n    private _pipelineCache: LetterboxPipeline | null = null;\n\n    /**\n     * Run the model once on zero-filled inputs, paying one-time costs up front.\n     *\n     * Worth more here than on a single-stage task: a fused pipeline is two models\n     * plus the bridge in one graph, so the first inference compiles shaders for\n     * all of it. Calling this while a loading spinner is still up moves that cost\n     * somewhere the user is already waiting.\n     *\n     * @param runs How many warm-up inferences to run. One is enough for WASM;\n     *   WebGPU sometimes settles on the second.\n     */\n    async warmup(runs: number = 1): Promise<void> {\n        const [width, height] = this._spec.inputSize;\n        for (let i = 0; i < runs; i++) {\n            const feeds: Record<string, ort.Tensor> = {\n                [INPUT_IMAGE]: toFloat32Tensor(zeroTensorData(width, height), [\n                    1,\n                    3,\n                    height,\n                    width,\n                ]),\n            };\n            if (this._spec.needsSourceImage) {\n                feeds[INPUT_SOURCE] = toFloat32Tensor(zeroTensorData(width, height), [\n                    1,\n                    3,\n                    height,\n                    width,\n                ]);\n                feeds[INPUT_SCALE] = toFloat32Tensor(new Float32Array([1]), [1]);\n                feeds[INPUT_PAD] = toFloat32Tensor(new Float32Array([0, 0]), [2]);\n            }\n            await this._session.run(feeds);\n        }\n    }\n\n    /**\n     * The fused preprocessing pipeline, built on first use.\n     *\n     * Lazily, because constructing it allocates canvases: a pipeline built in an\n     * environment without a canvas implementation stays constructible, and only\n     * fails if it is actually asked to preprocess something.\n     */\n    private get _pipeline(): LetterboxPipeline {\n        if (this._pipelineCache === null) {\n            const [width, height] = this._spec.inputSize;\n            this._pipelineCache = new LetterboxPipeline(width, height);\n        }\n        return this._pipelineCache;\n    }\n\n    /**\n     * Load a fused pipeline and resolve both label spaces.\n     *\n     * @param model The fused `.onnx` — a URL, an `ArrayBuffer`, or bytes.\n     * @param options Label overrides plus the usual session options.\n     * @throws {@link FusionError} when the model carries no pipeline metadata,\n     *   i.e. it is a plain detector or classifier rather than something\n     *   `ort_vision_sdk.compose` produced.\n     */\n    static async create(\n        model: ModelSource,\n        options: DetectClassifyOptions = {},\n    ): Promise<DetectClassify> {\n        const session = await OrtSession.create(model, options);\n        const spec = readFusionSpec(session.metadata);\n        if (spec === null) {\n            throw new FusionError(\n                \"This model carries no fused-pipeline metadata, so DetectClassify cannot tell how to \" +\n                    \"drive it. Build one with ort_vision_sdk.compose.fuse_detect_classify, or load a \" +\n                    \"plain model with Detector/Classifier instead.\",\n            );\n        }\n\n        const labels = resolveLabels(options.labels ?? spec.detectorNames ?? \"coco\");\n        const classifierLabels = resolveLabels(options.classifierLabels ?? spec.classifierNames, {\n            numClasses: classifierClasses(session) ?? undefined,\n        });\n        return new DetectClassify(\n            session,\n            spec,\n            labels,\n            indexNames(labels),\n            classifierLabels,\n            indexNames(classifierLabels),\n            options.raiseOnEmpty ?? false,\n        );\n    }\n\n    /** The pipeline configuration recorded in the model at fusion time. */\n    get spec(): FusionSpec {\n        return this._spec;\n    }\n\n    /** The `[width, height]` the detection stage runs at. */\n    get inputSize(): readonly [number, number] {\n        return this._spec.inputSize;\n    }\n\n    /** Detector class labels indexed by class id. */\n    get labels(): readonly string[] {\n        return this._labels;\n    }\n\n    /** Detector class id → class name (matches Ultralytics' `model.names`). */\n    get names(): Readonly<Record<number, string>> {\n        return this._names;\n    }\n\n    /** Classifier class labels indexed by class id. */\n    get classifierLabels(): readonly string[] {\n        return this._classifierLabels;\n    }\n\n    /** Classifier class id → class name. */\n    get classifierNames(): Readonly<Record<number, string>> {\n        return this._classifierNames;\n    }\n\n    /**\n     * Alias for {@link predict} — call the pipeline like a torch `nn.Module`.\n     *\n     * Use as `pipeline.call(img)` since JavaScript class instances are not\n     * callable; for direct invocation, prefer `pipeline.predict(img)`.\n     */\n    async call(\n        image: ImageInput,\n        options: DetectClassifyPredictOptions = {},\n    ): Promise<DetectClassifyResults[]> {\n        return this.predict(image, options);\n    }\n\n    /**\n     * Run the pipeline on a single image.\n     *\n     * The returned envelope carries a {@link Speed} breakdown in `speed`. Its\n     * `inference` figure covers detection *and* classification, since the\n     * pipeline runs them as one graph and no boundary between them is observable\n     * from outside.\n     */\n    async predict(\n        image: ImageInput,\n        options: DetectClassifyPredictOptions = {},\n    ): Promise<DetectClassifyResults[]> {\n        const timer = new SpeedTimer();\n        const path = typeof image === \"string\" ? image : null;\n        const original = await loadImage(image);\n        timer.stage(\"load\");\n        const { feeds, scale, padLeft, padTop } = this._preprocess(original);\n        timer.stage(\"preprocess\");\n        const outputs = await this._session.run(feeds);\n        this._pipeline.release();\n        timer.stage(\"inference\");\n\n        const probsTensor = output(outputs, OUTPUT_PROBS);\n        const boxes = floats(outputs, OUTPUT_BOXES);\n        const scores = floats(outputs, OUTPUT_SCORES);\n        const classes = integers(outputs, OUTPUT_CLASSES);\n        const probs = probsTensor.data as Float32Array;\n        const reported = integers(outputs, OUTPUT_NUM_DETECTIONS)[0] ?? 0;\n        const rows = Math.min(reported, Math.floor(boxes.length / 4));\n        const classCount = probsTensor.dims[probsTensor.dims.length - 1] ?? 0;\n\n        const allowed = options.classes === undefined ? null : new Set(options.classes);\n        const floor = options.confThreshold ?? 0;\n        const detections: DetectionResult[] = [];\n        for (let row = 0; row < rows; row++) {\n            const classId = classes[row] ?? 0;\n            const confidence = scores[row] ?? 0;\n            if (confidence < floor || (allowed !== null && !allowed.has(classId))) continue;\n\n            const bbox = this._toOriginal(boxes, row, { scale, padLeft, padTop, original });\n            const cropped = crop(original, bbox);\n            detections.push(\n                detection(\n                    classId,\n                    this._names[classId] ?? `class_${classId}`,\n                    confidence,\n                    bbox,\n                    cropped,\n                    this._classify(\n                        probs.subarray(row * classCount, (row + 1) * classCount),\n                        cropped,\n                        options.topK,\n                    ),\n                ),\n            );\n        }\n\n        requireDetections(detections.length, {\n            raiseOnEmpty: options.raiseOnEmpty ?? this._raiseOnEmpty,\n            confThreshold: Math.max(floor, this._spec.confThreshold),\n            classes: options.classes,\n            path,\n        });\n\n        const origShape: readonly [number, number] = [original.height, original.width];\n        timer.stage(\"postprocess\");\n        return [\n            new DetectClassifyResults(\n                bulkBoxes(detections, origShape),\n                detections,\n                this._names,\n                this._classifierNames,\n                original,\n                origShape,\n                path,\n                timer.speed(),\n            ),\n        ];\n    }\n\n    /**\n     * Letterbox the image and build the graph's feeds.\n     *\n     * The detector input runs through {@link LetterboxPipeline}, which fuses the\n     * resize, the padding and the HWC-to-CHW float conversion into one\n     * `drawImage` plus one readback, and reuses its output buffer between frames.\n     * That buffer goes straight to ONNX Runtime, so `_pipeline.release()` must not\n     * be called until the run resolves.\n     *\n     * A pipeline fused with `cropSource: \"original\"` also takes the untouched\n     * image as a second input, plus the scale and padding of the letterbox — that\n     * is what lets the graph undo the letterbox transform internally and crop at\n     * native resolution instead of from the downscaled copy. That one is **not**\n     * letterboxed by definition, so it does not go through the fused path.\n     */\n    private _preprocess(image: RGBImage): {\n        feeds: Record<string, ort.Tensor>;\n        scale: number;\n        padLeft: number;\n        padTop: number;\n    } {\n        const [width, height] = this._spec.inputSize;\n        const boxed = this._pipeline.run(image);\n        const feeds: Record<string, ort.Tensor> = {\n            [INPUT_IMAGE]: toFloat32Tensor(boxed.data, [1, 3, height, width]),\n        };\n\n        if (this._spec.needsSourceImage) {\n            feeds[INPUT_SOURCE] = tensorOf(image);\n            feeds[INPUT_SCALE] = toFloat32Tensor(new Float32Array([boxed.scale]), [1]);\n            feeds[INPUT_PAD] = toFloat32Tensor(\n                new Float32Array([boxed.padLeft, boxed.padTop]),\n                [2],\n            );\n        }\n        return { feeds, scale: boxed.scale, padLeft: boxed.padLeft, padTop: boxed.padTop };\n    }\n\n    /**\n     * Map one letterboxed xyxy row back onto the original image.\n     *\n     * The graph always reports boxes in the detector's letterboxed pixel space,\n     * whichever crop source it was fused with, so both sources agree here.\n     */\n    private _toOriginal(\n        boxes: Float32Array,\n        row: number,\n        context: {\n            scale: number;\n            padLeft: number;\n            padTop: number;\n            original: RGBImage;\n        },\n    ): BoundingBox {\n        const { scale, padLeft, padTop, original } = context;\n        const at = (offset: number): number => boxes[row * 4 + offset] ?? 0;\n        const clampX = (value: number): number => Math.min(Math.max(value, 0), original.width);\n        const clampY = (value: number): number => Math.min(Math.max(value, 0), original.height);\n        return new BoundingBox(\n            clampX((at(0) - padLeft) / scale),\n            clampY((at(1) - padTop) / scale),\n            clampX((at(2) - padLeft) / scale),\n            clampY((at(3) - padTop) / scale),\n        );\n    }\n\n    /**\n     * Turn one row of the classifier output into a result object.\n     *\n     * @param row The output row for this detection.\n     * @param image The crop the row describes, carried so callers can display\n     *   what was classified.\n     * @param k Optional truncation of the probability list.\n     */\n    private _classify(\n        row: Float32Array,\n        image: RGBImage,\n        k: number | undefined,\n    ): ClassificationResult {\n        const scores = this._spec.applySoftmax ? softmax(row) : row;\n        const { indices, values } = topK(scores, k ?? null);\n        const probabilities: ClassProbability[] = [];\n        for (let i = 0; i < indices.length; i++) {\n            const classId = indices[i] ?? 0;\n            const probability = values[i] ?? 0;\n            const className = this._classifierLabels[classId] ?? `class_${classId}`;\n            probabilities.push({\n                classId,\n                className,\n                probability,\n                cls: classId,\n                name: className,\n                conf: probability,\n            });\n        }\n        const top = probabilities[0] ?? {\n            classId: 0,\n            className: \"class_0\",\n            probability: 0,\n            cls: 0,\n            name: \"class_0\",\n            conf: 0,\n        };\n        return {\n            classId: top.classId,\n            className: top.className,\n            confidence: top.probability,\n            cls: top.classId,\n            name: top.className,\n            conf: top.probability,\n            image,\n            probabilities,\n        };\n    }\n}\n\n/**\n * Read the classifier stage's class count off the `probs` output shape.\n *\n * @param session The loaded pipeline session.\n * @returns The class count, or `null` when the graph leaves that axis dynamic\n *   or declares no `probs` output — in which case label resolution falls back\n *   to whatever the fusion recorded.\n */\nfunction classifierClasses(session: OrtSession): number | null {\n    const index = session.outputNames.indexOf(OUTPUT_PROBS);\n    if (index < 0) return null;\n    const shape = session.outputShapes[index];\n    if (shape === undefined || shape.length === 0) return null;\n    return shape[shape.length - 1] ?? null;\n}\n\n/**\n * Build a class id → name record from an ordered label list.\n *\n * @param labels Labels indexed by class id.\n * @returns The equivalent record.\n */\nfunction indexNames(labels: readonly string[]): Readonly<Record<number, string>> {\n    const names: Record<number, string> = {};\n    for (let i = 0; i < labels.length; i++) names[i] = labels[i] as string;\n    return names;\n}\n\n/**\n * Fetch an output tensor by name.\n *\n * @param outputs The run's results.\n * @param name The output's name in the pipeline contract.\n * @returns The tensor.\n * @throws {@link FusionError} when the graph does not carry that output, which\n *   means the file is not a pipeline this version can drive.\n */\nfunction output(outputs: Record<string, ort.Tensor>, name: string): ort.Tensor {\n    const tensor = outputs[name];\n    if (tensor === undefined) {\n        throw new FusionError(`The fused pipeline is missing its '${name}' output.`);\n    }\n    return tensor;\n}\n\n/**\n * Fetch a float output by name.\n *\n * @param outputs The run's results.\n * @param name The output's name in the pipeline contract.\n * @returns The output's data.\n * @throws {@link FusionError} when the graph does not carry that output.\n */\nfunction floats(outputs: Record<string, ort.Tensor>, name: string): Float32Array {\n    return output(outputs, name).data as Float32Array;\n}\n\n/**\n * Fetch an integer output by name, normalizing ORT's 64-bit representation.\n *\n * ONNX Runtime Web returns `int64` tensors as `BigInt64Array`, whose values do\n * not compare or index like numbers. Class ids and detection counts are always\n * small, so widening them to `number` here is lossless and keeps every caller\n * free of `BigInt` handling.\n *\n * @param outputs The run's results.\n * @param name The output's name in the pipeline contract.\n * @returns The output's values as plain numbers.\n * @throws {@link FusionError} when the graph does not carry that output.\n */\nfunction integers(outputs: Record<string, ort.Tensor>, name: string): number[] {\n    const data = output(outputs, name).data as BigInt64Array | Int32Array | Float32Array;\n    const values: number[] = [];\n    for (let i = 0; i < data.length; i++) values.push(Number(data[i]));\n    return values;\n}\n\n/**\n * Convert an image to the `[1, 3, H, W]` float32 tensor the graph expects.\n *\n * @param image The image to convert.\n * @returns The batched CHW tensor, scaled to `[0, 1]`.\n */\nfunction tensorOf(image: RGBImage): ort.Tensor {\n    const chw = toCHW(toFloat32(image), image.width, image.height, 3);\n    return toFloat32Tensor(chw, [1, 3, image.height, image.width]);\n}\n\n/**\n * Cut the box region out of the original image.\n *\n * @param image The source image.\n * @param bbox The box, in original-image pixel coordinates.\n * @returns The cropped region, or a zero-sized image for a box with no area.\n */\nfunction crop(image: RGBImage, bbox: BoundingBox): RGBImage {\n    const [rawX1, rawY1, rawX2, rawY2] = bbox.asIntXyxy();\n    const x1 = Math.max(0, rawX1);\n    const y1 = Math.max(0, rawY1);\n    const x2 = Math.min(image.width, rawX2);\n    const y2 = Math.min(image.height, rawY2);\n    if (x2 <= x1 || y2 <= y1) return new RGBImage(new Uint8Array(0), 0, 0);\n\n    const width = x2 - x1;\n    const height = y2 - y1;\n    const out = new Uint8Array(width * height * 3);\n    for (let row = 0; row < height; row++) {\n        const offset = ((y1 + row) * image.width + x1) * 3;\n        out.set(image.data.subarray(offset, offset + width * 3), row * width * 3);\n    }\n    return new RGBImage(out, width, height);\n}\n\n/**\n * Assemble one detection, filling the Ultralytics-style aliases.\n *\n * @param classId Detector class index.\n * @param className Detector class name.\n * @param confidence Detection score.\n * @param bbox Box in original-image coordinates.\n * @param croppedImage The region the box covers.\n * @param classification What the classification stage said about that region.\n * @returns The detection object.\n */\nfunction detection(\n    classId: number,\n    className: string,\n    confidence: number,\n    bbox: BoundingBox,\n    croppedImage: RGBImage,\n    classification: ClassificationResult,\n): DetectionResult {\n    return {\n        classId,\n        className,\n        confidence,\n        bbox,\n        cls: classId,\n        name: className,\n        conf: confidence,\n        box: bbox,\n        croppedImage,\n        classification,\n    };\n}\n\n/**\n * Assemble the bulk-array `Boxes` view from per-instance detections.\n *\n * @param detections The surviving detections.\n * @param origShape `[height, width]` of the original image.\n * @returns The bulk view, empty when nothing survived.\n */\nfunction bulkBoxes(\n    detections: readonly DetectionResult[],\n    origShape: readonly [number, number],\n): Boxes {\n    const xyxy = new Float32Array(detections.length * 4);\n    const cls = new Int32Array(detections.length);\n    const conf = new Float32Array(detections.length);\n    detections.forEach((entry, index) => {\n        const [x1, y1, x2, y2] = entry.bbox.xyxy;\n        xyxy[index * 4] = x1;\n        xyxy[index * 4 + 1] = y1;\n        xyxy[index * 4 + 2] = x2;\n        xyxy[index * 4 + 3] = y2;\n        cls[index] = entry.classId;\n        conf[index] = entry.confidence;\n    });\n    return new Boxes(xyxy, cls, conf, origShape);\n}\n"],"mappings":"wYAqGA,IAAa,EAAb,MAAa,UAAuB,EAAA,UAAW,CAGtB,MACA,QACA,OACA,kBACA,iBACA,cAPrB,YACI,EACA,EACA,EACA,EACA,EACA,EACA,EACF,CACE,MAAM,CAAO,EAPI,KAAA,MAAA,EACA,KAAA,QAAA,EACA,KAAA,OAAA,EACA,KAAA,kBAAA,EACA,KAAA,iBAAA,EACA,KAAA,cAAA,CAGrB,CAEA,eAAmD,KAanD,MAAM,OAAO,EAAe,EAAkB,CAC1C,GAAM,CAAC,EAAO,GAAU,KAAK,MAAM,UACnC,IAAK,IAAI,EAAI,EAAG,EAAI,EAAM,IAAK,CAC3B,IAAM,EAAoC,EACrC,EAAA,aAAc,EAAA,gBAAgB,EAAA,eAAe,EAAO,CAAM,EAAG,CAC1D,EACA,EACA,EACA,CACJ,CAAC,CACL,EACI,KAAK,MAAM,mBACX,EAAM,EAAA,cAAgB,EAAA,gBAAgB,EAAA,eAAe,EAAO,CAAM,EAAG,CACjE,EACA,EACA,EACA,CACJ,CAAC,EACD,EAAM,EAAA,aAAe,EAAA,gBAAgB,IAAI,aAAa,CAAC,CAAC,CAAC,EAAG,CAAC,CAAC,CAAC,EAC/D,EAAM,EAAA,WAAa,EAAA,gBAAgB,IAAI,aAAa,CAAC,EAAG,CAAC,CAAC,EAAG,CAAC,CAAC,CAAC,GAEpE,MAAM,KAAK,SAAS,IAAI,CAAK,CACjC,CACJ,CASA,IAAY,WAA+B,CACvC,GAAI,KAAK,iBAAmB,KAAM,CAC9B,GAAM,CAAC,EAAO,GAAU,KAAK,MAAM,UACnC,KAAK,eAAiB,IAAI,EAAA,kBAAkB,EAAO,CAAM,CAC7D,CACA,OAAO,KAAK,cAChB,CAWA,aAAa,OACT,EACA,EAAiC,CAAC,EACX,CACvB,IAAM,EAAU,MAAM,EAAA,WAAW,OAAO,EAAO,CAAO,EAChD,EAAO,EAAA,eAAe,EAAQ,QAAQ,EAC5C,GAAI,IAAS,KACT,MAAM,IAAI,EAAA,YACN,mNAGJ,EAGJ,IAAM,EAAS,EAAA,cAAc,EAAQ,QAAU,EAAK,eAAiB,MAAM,EACrE,EAAmB,EAAA,cAAc,EAAQ,kBAAoB,EAAK,gBAAiB,CACrF,WAAY,EAAkB,CAAO,GAAK,IAAA,EAC9C,CAAC,EACD,OAAO,IAAI,EACP,EACA,EACA,EACA,EAAW,CAAM,EACjB,EACA,EAAW,CAAgB,EAC3B,EAAQ,cAAgB,EAC5B,CACJ,CAGA,IAAI,MAAmB,CACnB,OAAO,KAAK,KAChB,CAGA,IAAI,WAAuC,CACvC,OAAO,KAAK,MAAM,SACtB,CAGA,IAAI,QAA4B,CAC5B,OAAO,KAAK,OAChB,CAGA,IAAI,OAA0C,CAC1C,OAAO,KAAK,MAChB,CAGA,IAAI,kBAAsC,CACtC,OAAO,KAAK,iBAChB,CAGA,IAAI,iBAAoD,CACpD,OAAO,KAAK,gBAChB,CAQA,MAAM,KACF,EACA,EAAwC,CAAC,EACT,CAChC,OAAO,KAAK,QAAQ,EAAO,CAAO,CACtC,CAUA,MAAM,QACF,EACA,EAAwC,CAAC,EACT,CAChC,IAAM,EAAQ,IAAI,EAAA,WACZ,EAAO,OAAO,GAAU,SAAW,EAAQ,KAC3C,EAAW,MAAM,EAAA,UAAU,CAAK,EACtC,EAAM,MAAM,MAAM,EAClB,GAAM,CAAE,QAAO,QAAO,UAAS,UAAW,KAAK,YAAY,CAAQ,EACnE,EAAM,MAAM,YAAY,EACxB,IAAM,EAAU,MAAM,KAAK,SAAS,IAAI,CAAK,EAC7C,KAAK,UAAU,QAAQ,EACvB,EAAM,MAAM,WAAW,EAEvB,IAAM,EAAc,EAAO,EAAS,EAAA,YAAY,EAC1C,EAAQ,EAAO,EAAS,EAAA,YAAY,EACpC,EAAS,EAAO,EAAS,EAAA,aAAa,EACtC,EAAU,EAAS,EAAS,EAAA,cAAc,EAC1C,EAAQ,EAAY,KACpB,EAAW,EAAS,EAAA,gBAA8B,CAAC,CAAC,IAAM,EAC1D,EAAO,KAAK,IAAI,EAAU,KAAK,MAAM,EAAM,OAAS,CAAC,CAAC,EACtD,EAAa,EAAY,KAAK,EAAY,KAAK,OAAS,IAAM,EAE9D,EAAU,EAAQ,UAAY,IAAA,GAAY,KAAO,IAAI,IAAI,EAAQ,OAAO,EACxE,EAAQ,EAAQ,eAAiB,EACjC,EAAgC,CAAC,EACvC,IAAK,IAAI,EAAM,EAAG,EAAM,EAAM,IAAO,CACjC,IAAM,EAAU,EAAQ,IAAQ,EAC1B,EAAa,EAAO,IAAQ,EAClC,GAAI,EAAa,GAAU,IAAY,MAAQ,CAAC,EAAQ,IAAI,CAAO,EAAI,SAEvE,IAAM,EAAO,KAAK,YAAY,EAAO,EAAK,CAAE,QAAO,UAAS,SAAQ,UAAS,CAAC,EACxE,EAAU,EAAK,EAAU,CAAI,EACnC,EAAW,KACP,EACI,EACA,KAAK,OAAO,IAAY,SAAS,IACjC,EACA,EACA,EACA,KAAK,UACD,EAAM,SAAS,EAAM,GAAa,EAAM,GAAK,CAAU,EACvD,EACA,EAAQ,IACZ,CACJ,CACJ,CACJ,CAEA,EAAA,kBAAkB,EAAW,OAAQ,CACjC,aAAc,EAAQ,cAAgB,KAAK,cAC3C,cAAe,KAAK,IAAI,EAAO,KAAK,MAAM,aAAa,EACvD,QAAS,EAAQ,QACjB,MACJ,CAAC,EAED,IAAM,EAAuC,CAAC,EAAS,OAAQ,EAAS,KAAK,EAE7E,OADA,EAAM,MAAM,aAAa,EAClB,CACH,IAAI,EAAA,sBACA,EAAU,EAAY,CAAS,EAC/B,EACA,KAAK,OACL,KAAK,iBACL,EACA,EACA,EACA,EAAM,MAAM,CAChB,CACJ,CACJ,CAiBA,YAAoB,EAKlB,CACE,GAAM,CAAC,EAAO,GAAU,KAAK,MAAM,UAC7B,EAAQ,KAAK,UAAU,IAAI,CAAK,EAChC,EAAoC,EACrC,EAAA,aAAc,EAAA,gBAAgB,EAAM,KAAM,CAAC,EAAG,EAAG,EAAQ,CAAK,CAAC,CACpE,EAUA,OARI,KAAK,MAAM,mBACX,EAAM,EAAA,cAAgB,EAAS,CAAK,EACpC,EAAM,EAAA,aAAe,EAAA,gBAAgB,IAAI,aAAa,CAAC,EAAM,KAAK,CAAC,EAAG,CAAC,CAAC,CAAC,EACzE,EAAM,EAAA,WAAa,EAAA,gBACf,IAAI,aAAa,CAAC,EAAM,QAAS,EAAM,MAAM,CAAC,EAC9C,CAAC,CAAC,CACN,GAEG,CAAE,QAAO,MAAO,EAAM,MAAO,QAAS,EAAM,QAAS,OAAQ,EAAM,MAAO,CACrF,CAQA,YACI,EACA,EACA,EAMW,CACX,GAAM,CAAE,QAAO,UAAS,SAAQ,YAAa,EACvC,EAAM,GAA2B,EAAM,EAAM,EAAI,IAAW,EAC5D,EAAU,GAA0B,KAAK,IAAI,KAAK,IAAI,EAAO,CAAC,EAAG,EAAS,KAAK,EAC/E,EAAU,GAA0B,KAAK,IAAI,KAAK,IAAI,EAAO,CAAC,EAAG,EAAS,MAAM,EACtF,OAAO,IAAI,EAAA,YACP,GAAQ,EAAG,CAAC,EAAI,GAAW,CAAK,EAChC,GAAQ,EAAG,CAAC,EAAI,GAAU,CAAK,EAC/B,GAAQ,EAAG,CAAC,EAAI,GAAW,CAAK,EAChC,GAAQ,EAAG,CAAC,EAAI,GAAU,CAAK,CACnC,CACJ,CAUA,UACI,EACA,EACA,EACoB,CACpB,IAAM,EAAS,KAAK,MAAM,aAAe,EAAA,QAAQ,CAAG,EAAI,EAClD,CAAE,UAAS,UAAW,EAAA,KAAK,EAAQ,GAAK,IAAI,EAC5C,EAAoC,CAAC,EAC3C,IAAK,IAAI,EAAI,EAAG,EAAI,EAAQ,OAAQ,IAAK,CACrC,IAAM,EAAU,EAAQ,IAAM,EACxB,EAAc,EAAO,IAAM,EAC3B,EAAY,KAAK,kBAAkB,IAAY,SAAS,IAC9D,EAAc,KAAK,CACf,UACA,YACA,cACA,IAAK,EACL,KAAM,EACN,KAAM,CACV,CAAC,CACL,CACA,IAAM,EAAM,EAAc,IAAM,CAC5B,QAAS,EACT,UAAW,UACX,YAAa,EACb,IAAK,EACL,KAAM,UACN,KAAM,CACV,EACA,MAAO,CACH,QAAS,EAAI,QACb,UAAW,EAAI,UACf,WAAY,EAAI,YAChB,IAAK,EAAI,QACT,KAAM,EAAI,UACV,KAAM,EAAI,YACV,QACA,eACJ,CACJ,CACJ,EAUA,SAAS,EAAkB,EAAoC,CAC3D,IAAM,EAAQ,EAAQ,YAAY,QAAQ,EAAA,YAAY,EACtD,GAAI,EAAQ,EAAG,OAAO,KACtB,IAAM,EAAQ,EAAQ,aAAa,GAEnC,OADI,IAAU,IAAA,IAAa,EAAM,SAAW,EAAU,KAC/C,EAAM,EAAM,OAAS,IAAM,IACtC,CAQA,SAAS,EAAW,EAA6D,CAC7E,IAAM,EAAgC,CAAC,EACvC,IAAK,IAAI,EAAI,EAAG,EAAI,EAAO,OAAQ,IAAK,EAAM,GAAK,EAAO,GAC1D,OAAO,CACX,CAWA,SAAS,EAAO,EAAqC,EAA0B,CAC3E,IAAM,EAAS,EAAQ,GACvB,GAAI,IAAW,IAAA,GACX,MAAM,IAAI,EAAA,YAAY,sCAAsC,EAAK,UAAU,EAE/E,OAAO,CACX,CAUA,SAAS,EAAO,EAAqC,EAA4B,CAC7E,OAAO,EAAO,EAAS,CAAI,CAAC,CAAC,IACjC,CAeA,SAAS,EAAS,EAAqC,EAAwB,CAC3E,IAAM,EAAO,EAAO,EAAS,CAAI,CAAC,CAAC,KAC7B,EAAmB,CAAC,EAC1B,IAAK,IAAI,EAAI,EAAG,EAAI,EAAK,OAAQ,IAAK,EAAO,KAAK,OAAO,EAAK,EAAE,CAAC,EACjE,OAAO,CACX,CAQA,SAAS,EAAS,EAA6B,CAC3C,IAAM,EAAM,EAAA,MAAM,EAAA,UAAU,CAAK,EAAG,EAAM,MAAO,EAAM,OAAQ,CAAC,EAChE,OAAO,EAAA,gBAAgB,EAAK,CAAC,EAAG,EAAG,EAAM,OAAQ,EAAM,KAAK,CAAC,CACjE,CASA,SAAS,EAAK,EAAiB,EAA6B,CACxD,GAAM,CAAC,EAAO,EAAO,EAAO,GAAS,EAAK,UAAU,EAC9C,EAAK,KAAK,IAAI,EAAG,CAAK,EACtB,EAAK,KAAK,IAAI,EAAG,CAAK,EACtB,EAAK,KAAK,IAAI,EAAM,MAAO,CAAK,EAChC,EAAK,KAAK,IAAI,EAAM,OAAQ,CAAK,EACvC,GAAI,GAAM,GAAM,GAAM,EAAI,OAAO,IAAI,EAAA,SAAS,IAAI,WAAe,EAAG,CAAC,EAErE,IAAM,EAAQ,EAAK,EACb,EAAS,EAAK,EACd,EAAM,IAAI,WAAW,EAAQ,EAAS,CAAC,EAC7C,IAAK,IAAI,EAAM,EAAG,EAAM,EAAQ,IAAO,CACnC,IAAM,IAAW,EAAK,GAAO,EAAM,MAAQ,GAAM,EACjD,EAAI,IAAI,EAAM,KAAK,SAAS,EAAQ,EAAS,EAAQ,CAAC,EAAG,EAAM,EAAQ,CAAC,CAC5E,CACA,OAAO,IAAI,EAAA,SAAS,EAAK,EAAO,CAAM,CAC1C,CAaA,SAAS,EACL,EACA,EACA,EACA,EACA,EACA,EACe,CACf,MAAO,CACH,UACA,YACA,aACA,OACA,IAAK,EACL,KAAM,EACN,KAAM,EACN,IAAK,EACL,eACA,gBACJ,CACJ,CASA,SAAS,EACL,EACA,EACK,CACL,IAAM,EAAO,IAAI,aAAa,EAAW,OAAS,CAAC,EAC7C,EAAM,IAAI,WAAW,EAAW,MAAM,EACtC,EAAO,IAAI,aAAa,EAAW,MAAM,EAU/C,OATA,EAAW,SAAS,EAAO,IAAU,CACjC,GAAM,CAAC,EAAI,EAAI,EAAI,GAAM,EAAM,KAAK,KACpC,EAAK,EAAQ,GAAK,EAClB,EAAK,EAAQ,EAAI,GAAK,EACtB,EAAK,EAAQ,EAAI,GAAK,EACtB,EAAK,EAAQ,EAAI,GAAK,EACtB,EAAI,GAAS,EAAM,QACnB,EAAK,GAAS,EAAM,UACxB,CAAC,EACM,IAAI,EAAA,MAAM,EAAM,EAAK,EAAM,CAAS,CAC/C"}
+{"version":3,"file":"detectClassify.cjs","names":[],"sources":["../../../src/vision/tasks/detectClassify.ts"],"sourcesContent":["/** @generated Vendored from @mauriciobenjamin700/ort-vision-sdk-web. Do not hand-edit — regenerate with `npm run vendor:vision`. */\n/**\n * Run a fused detect→classify pipeline in the browser.\n *\n * The file this loads was built by the Python SDK's `ort_vision_sdk.compose`,\n * and it already contains both models plus the crop-and-resize bridge between\n * them. That matters far more in a tab than on a server: two models mean two\n * `.onnx` downloads, two WASM/WebGPU session initializations, and a per-crop\n * round trip through JavaScript to slice, resize and restack the regions before\n * the second model can see them. A fused pipeline has one download, one session\n * and no round trip — the crops are produced and consumed inside the graph.\n */\n\nimport type * as ort from \"onnxruntime-web\";\n\nimport { FusionError } from \"../core/exceptions\";\nimport { type ModelSource, type OrtSessionOptions, OrtSession } from \"../core/session\";\nimport { SpeedTimer } from \"../core/timing\";\nimport {\n    INPUT_IMAGE,\n    INPUT_PAD,\n    INPUT_SCALE,\n    INPUT_SOURCE,\n    OUTPUT_BOXES,\n    OUTPUT_CLASSES,\n    OUTPUT_NUM_DETECTIONS,\n    OUTPUT_MASKS,\n    OUTPUT_PROBS,\n    OUTPUT_SCORES,\n    type FusionSpec,\n    readFusionSpec,\n} from \"../fusion\";\nimport { type ImageInput, loadImageSource } from \"../io/image\";\nimport { type LabelSpec, resolveLabels } from \"../labels\";\nimport { softmax, topK } from \"../postprocess/classification\";\nimport { asFloat32Array } from \"../core/dtypes\";\nimport { toCHW, toFloat32, toFloat32Tensor } from \"../preprocess/image\";\nimport { LetterboxPipeline, zeroTensorData } from \"../preprocess/pipeline\";\nimport { Boxes, DetectClassifyResults } from \"../results\";\nimport type { RGBImage } from \"../types\";\nimport {\n    BoundingBox,\n    Mask,\n    type ClassProbability,\n    type ClassificationResult,\n    type DetectionResult,\n} from \"../types\";\nimport { VisionTask, cropToBox, requireDetections } from \"./base\";\n\nexport interface DetectClassifyOptions extends OrtSessionOptions {\n    /**\n     * Class label spec for the **detection** stage — see {@link resolveLabels}.\n     * Defaults to the names recorded at fusion time, falling back to the COCO\n     * 80-class preset when the fusion recorded none.\n     */\n    readonly labels?: LabelSpec;\n    /**\n     * Class label spec for the **classification** stage. Defaults to the recorded\n     * names, falling back to generated `class_<id>` names.\n     */\n    readonly classifierLabels?: LabelSpec;\n    /**\n     * If `true`, a run that finds nothing throws {@link NoDetectionsError}\n     * instead of returning an empty envelope. Default `false`, because looking\n     * and finding nothing is a successful inference. Turn it on when an empty\n     * result means the surrounding pipeline should stop rather than carry on with\n     * zero rows. Can be overridden per `predict` call.\n     */\n    readonly raiseOnEmpty?: boolean;\n}\n\nexport interface DetectClassifyPredictOptions {\n    /**\n     * Drop detections scoring below this. The graph's own NMS threshold was fixed\n     * at fusion time and cannot be lowered here — this only filters further.\n     */\n    readonly confThreshold?: number;\n    /** If set, keep only detections whose detector `classId` is in this list. */\n    readonly classes?: readonly number[];\n    /** Truncate each detection's `classification.probabilities` to its top-k entries. */\n    readonly topK?: number;\n    /** Override the constructor's `raiseOnEmpty` setting for this call. */\n    readonly raiseOnEmpty?: boolean;\n}\n\n/**\n * Detector and classifier running as a single ONNX model.\n *\n * Everything the pipeline needs to know about itself — the resolution to\n * letterbox to, whether it wants the full-resolution image as well, whether its\n * classifier output still needs a softmax, the class names of both stages — was\n * written into the file at fusion time and is read back here. Nothing is\n * restated on the JavaScript side, so nothing can drift out of step with the\n * Python side that built it.\n *\n * @example\n * ```typescript\n * const pipeline = await DetectClassify.create(\"/models/pipeline.onnx\");\n * const result = (await pipeline.predict(\"/images/flock.jpg\"))[0];\n * for (const detection of result) {\n *   console.log(detection.name, detection.conf, detection.classification?.name);\n * }\n * ```\n */\nexport class DetectClassify extends VisionTask {\n    private constructor(\n        session: OrtSession,\n        private readonly _spec: FusionSpec,\n        private readonly _labels: readonly string[],\n        private readonly _names: Readonly<Record<number, string>>,\n        private readonly _classifierLabels: readonly string[],\n        private readonly _classifierNames: Readonly<Record<number, string>>,\n        private readonly _raiseOnEmpty: boolean,\n    ) {\n        super(session);\n    }\n\n    private _pipelineCache: LetterboxPipeline | null = null;\n\n    /**\n     * Run the model once on zero-filled inputs, paying one-time costs up front.\n     *\n     * Worth more here than on a single-stage task: a fused pipeline is two models\n     * plus the bridge in one graph, so the first inference compiles shaders for\n     * all of it. Calling this while a loading spinner is still up moves that cost\n     * somewhere the user is already waiting.\n     *\n     * @param runs How many warm-up inferences to run. One is enough for WASM;\n     *   WebGPU sometimes settles on the second.\n     */\n    async warmup(runs: number = 1): Promise<void> {\n        const [width, height] = this._spec.inputSize;\n        for (let i = 0; i < runs; i++) {\n            const feeds: Record<string, ort.Tensor> = {\n                [INPUT_IMAGE]: toFloat32Tensor(zeroTensorData(width, height), [\n                    1,\n                    3,\n                    height,\n                    width,\n                ]),\n            };\n            if (this._spec.needsSourceImage) {\n                feeds[INPUT_SOURCE] = toFloat32Tensor(zeroTensorData(width, height), [\n                    1,\n                    3,\n                    height,\n                    width,\n                ]);\n                feeds[INPUT_SCALE] = toFloat32Tensor(new Float32Array([1]), [1]);\n                feeds[INPUT_PAD] = toFloat32Tensor(new Float32Array([0, 0]), [2]);\n            }\n            await this._session.run(feeds);\n        }\n    }\n\n    /**\n     * The fused preprocessing pipeline, built on first use.\n     *\n     * Lazily, because constructing it allocates canvases: a pipeline built in an\n     * environment without a canvas implementation stays constructible, and only\n     * fails if it is actually asked to preprocess something.\n     */\n    private get _pipeline(): LetterboxPipeline {\n        if (this._pipelineCache === null) {\n            const [width, height] = this._spec.inputSize;\n            this._pipelineCache = new LetterboxPipeline(width, height);\n        }\n        return this._pipelineCache;\n    }\n\n    /**\n     * Load a fused pipeline and resolve both label spaces.\n     *\n     * @param model The fused `.onnx` — a URL, an `ArrayBuffer`, or bytes.\n     * @param options Label overrides plus the usual session options.\n     * @throws {@link FusionError} when the model carries no pipeline metadata,\n     *   i.e. it is a plain detector or classifier rather than something\n     *   `ort_vision_sdk.compose` produced.\n     */\n    static async create(\n        model: ModelSource,\n        options: DetectClassifyOptions = {},\n    ): Promise<DetectClassify> {\n        const session = await OrtSession.create(model, options);\n        const spec = readFusionSpec(session.metadata);\n        if (spec === null) {\n            throw new FusionError(\n                \"This model carries no fused-pipeline metadata, so DetectClassify cannot tell how to \" +\n                    \"drive it. Build one with ort_vision_sdk.compose.fuse_detect_classify, or load a \" +\n                    \"plain model with Detector/Classifier instead.\",\n            );\n        }\n\n        const labels = resolveLabels(options.labels ?? spec.detectorNames ?? \"coco\");\n        const classifierLabels = resolveLabels(options.classifierLabels ?? spec.classifierNames, {\n            numClasses: classifierClasses(session) ?? undefined,\n        });\n        return new DetectClassify(\n            session,\n            spec,\n            labels,\n            indexNames(labels),\n            classifierLabels,\n            indexNames(classifierLabels),\n            options.raiseOnEmpty ?? false,\n        );\n    }\n\n    /** The pipeline configuration recorded in the model at fusion time. */\n    get spec(): FusionSpec {\n        return this._spec;\n    }\n\n    /** The `[width, height]` the detection stage runs at. */\n    get inputSize(): readonly [number, number] {\n        return this._spec.inputSize;\n    }\n\n    /** Detector class labels indexed by class id. */\n    get labels(): readonly string[] {\n        return this._labels;\n    }\n\n    /** Detector class id → class name (matches Ultralytics' `model.names`). */\n    get names(): Readonly<Record<number, string>> {\n        return this._names;\n    }\n\n    /** Classifier class labels indexed by class id. */\n    get classifierLabels(): readonly string[] {\n        return this._classifierLabels;\n    }\n\n    /** Classifier class id → class name. */\n    get classifierNames(): Readonly<Record<number, string>> {\n        return this._classifierNames;\n    }\n\n    /**\n     * Alias for {@link predict} — call the pipeline like a torch `nn.Module`.\n     *\n     * Use as `pipeline.call(img)` since JavaScript class instances are not\n     * callable; for direct invocation, prefer `pipeline.predict(img)`.\n     */\n    async call(\n        image: ImageInput,\n        options: DetectClassifyPredictOptions = {},\n    ): Promise<DetectClassifyResults[]> {\n        return this.predict(image, options);\n    }\n\n    /**\n     * Run the pipeline on a single image.\n     *\n     * The returned envelope carries a {@link Speed} breakdown in `speed`. Its\n     * `inference` figure covers detection *and* classification, since the\n     * pipeline runs them as one graph and no boundary between them is observable\n     * from outside.\n     */\n    async predict(\n        image: ImageInput,\n        options: DetectClassifyPredictOptions = {},\n    ): Promise<DetectClassifyResults[]> {\n        const timer = new SpeedTimer();\n        const path = typeof image === \"string\" ? image : null;\n        const { image: original, canvas } = await loadImageSource(image);\n        timer.stage(\"load\");\n        const { feeds, scale, padLeft, padTop, reused } = this._preprocess(original, canvas);\n        timer.stage(\"preprocess\");\n        let outputs: Record<string, ort.Tensor>;\n        try {\n            outputs = await this._session.run(feeds);\n        } finally {\n            this._pipeline.release({ reused });\n        }\n        timer.stage(\"inference\");\n\n        const probsTensor = output(outputs, OUTPUT_PROBS);\n        const boxes = floats(outputs, OUTPUT_BOXES);\n        const scores = floats(outputs, OUTPUT_SCORES);\n        const classes = integers(outputs, OUTPUT_CLASSES);\n        const probs = asFloat32Array(probsTensor.data);\n        const reported = integers(outputs, OUTPUT_NUM_DETECTIONS)[0] ?? 0;\n        const masks = this._spec.hasMasks ? floats(outputs, OUTPUT_MASKS) : null;\n        const [cropWidth, cropHeight] = this._spec.cropSize;\n        const rows = Math.min(reported, Math.floor(boxes.length / 4));\n        const classCount = probsTensor.dims[probsTensor.dims.length - 1] ?? 0;\n\n        const allowed = options.classes === undefined ? null : new Set(options.classes);\n        const floor = options.confThreshold ?? 0;\n        const detections: DetectionResult[] = [];\n        for (let row = 0; row < rows; row++) {\n            const classId = classes[row] ?? 0;\n            const confidence = scores[row] ?? 0;\n            if (confidence < floor || (allowed !== null && !allowed.has(classId))) continue;\n\n            const bbox = this._toOriginal(boxes, row, { scale, padLeft, padTop, original });\n            const cropped = cropToBox(original, bbox);\n            detections.push(\n                detection(\n                    classId,\n                    this._names[classId] ?? `class_${classId}`,\n                    confidence,\n                    bbox,\n                    cropped,\n                    this._classify(\n                        probs.subarray(row * classCount, (row + 1) * classCount),\n                        cropped,\n                        options.topK,\n                    ),\n                    masks === null ? null : maskToBox(masks, row, cropWidth, cropHeight, cropped),\n                ),\n            );\n        }\n\n        requireDetections(detections.length, {\n            raiseOnEmpty: options.raiseOnEmpty ?? this._raiseOnEmpty,\n            confThreshold: Math.max(floor, this._spec.confThreshold),\n            classes: options.classes,\n            path,\n        });\n\n        const origShape: readonly [number, number] = [original.height, original.width];\n        timer.stage(\"postprocess\");\n        return [\n            new DetectClassifyResults(\n                bulkBoxes(detections, origShape),\n                detections,\n                this._names,\n                this._classifierNames,\n                original,\n                origShape,\n                path,\n                timer.speed(),\n            ),\n        ];\n    }\n\n    /**\n     * Letterbox the image and build the graph's feeds.\n     *\n     * The detector input runs through {@link LetterboxPipeline}, which fuses the\n     * resize, the padding and the HWC-to-CHW float conversion into one\n     * `drawImage` plus one readback, and reuses its output buffer between frames.\n     * That buffer goes straight to ONNX Runtime, so `_pipeline.release()` must not\n     * be called until the run resolves.\n     *\n     * A pipeline fused with `cropSource: \"original\"` also takes the untouched\n     * image as a second input, plus the scale and padding of the letterbox — that\n     * is what lets the graph undo the letterbox transform internally and crop at\n     * native resolution instead of from the downscaled copy. That one is **not**\n     * letterboxed by definition, so it does not go through the fused path.\n     *\n     * @param image The decoded input.\n     * @param canvas The opaque canvas it was decoded on, when there is one — see\n     *   {@link LetterboxPipeline.run}.\n     */\n    private _preprocess(\n        image: RGBImage,\n        canvas: CanvasImageSource | null,\n    ): {\n        feeds: Record<string, ort.Tensor>;\n        scale: number;\n        padLeft: number;\n        padTop: number;\n        reused: boolean;\n    } {\n        const [width, height] = this._spec.inputSize;\n        const boxed = this._pipeline.run(image, canvas);\n        const feeds: Record<string, ort.Tensor> = {\n            [INPUT_IMAGE]: toFloat32Tensor(boxed.data, [1, 3, height, width]),\n        };\n\n        if (this._spec.needsSourceImage) {\n            feeds[INPUT_SOURCE] = tensorOf(image);\n            feeds[INPUT_SCALE] = toFloat32Tensor(new Float32Array([boxed.scale]), [1]);\n            feeds[INPUT_PAD] = toFloat32Tensor(\n                new Float32Array([boxed.padLeft, boxed.padTop]),\n                [2],\n            );\n        }\n        return {\n            feeds,\n            scale: boxed.scale,\n            padLeft: boxed.padLeft,\n            padTop: boxed.padTop,\n            reused: boxed.reused,\n        };\n    }\n\n    /**\n     * Map one letterboxed xyxy row back onto the original image.\n     *\n     * The graph always reports boxes in the detector's letterboxed pixel space,\n     * whichever crop source it was fused with, so both sources agree here.\n     */\n    private _toOriginal(\n        boxes: Float32Array,\n        row: number,\n        context: {\n            scale: number;\n            padLeft: number;\n            padTop: number;\n            original: RGBImage;\n        },\n    ): BoundingBox {\n        const { scale, padLeft, padTop, original } = context;\n        const at = (offset: number): number => boxes[row * 4 + offset] ?? 0;\n        const clampX = (value: number): number => Math.min(Math.max(value, 0), original.width);\n        const clampY = (value: number): number => Math.min(Math.max(value, 0), original.height);\n        return new BoundingBox(\n            clampX((at(0) - padLeft) / scale),\n            clampY((at(1) - padTop) / scale),\n            clampX((at(2) - padLeft) / scale),\n            clampY((at(3) - padTop) / scale),\n        );\n    }\n\n    /**\n     * Turn one row of the classifier output into a result object.\n     *\n     * @param row The output row for this detection.\n     * @param image The crop the row describes, carried so callers can display\n     *   what was classified.\n     * @param k Optional truncation of the probability list.\n     */\n    private _classify(\n        row: Float32Array,\n        image: RGBImage,\n        k: number | undefined,\n    ): ClassificationResult {\n        const scores = this._spec.applySoftmax ? softmax(row) : row;\n        const { indices, values } = topK(scores, k ?? null);\n        const probabilities: ClassProbability[] = [];\n        for (let i = 0; i < indices.length; i++) {\n            const classId = indices[i] ?? 0;\n            const probability = values[i] ?? 0;\n            const className = this._classifierLabels[classId] ?? `class_${classId}`;\n            probabilities.push({\n                classId,\n                className,\n                probability,\n                cls: classId,\n                name: className,\n                conf: probability,\n            });\n        }\n        const top = probabilities[0] ?? {\n            classId: 0,\n            className: \"class_0\",\n            probability: 0,\n            cls: 0,\n            name: \"class_0\",\n            conf: 0,\n        };\n        return {\n            classId: top.classId,\n            className: top.className,\n            confidence: top.probability,\n            cls: top.classId,\n            name: top.className,\n            conf: top.probability,\n            image,\n            probabilities,\n        };\n    }\n}\n\n/**\n * Read the classifier stage's class count off the `probs` output shape.\n *\n * @param session The loaded pipeline session.\n * @returns The class count, or `null` when the graph leaves that axis dynamic\n *   or declares no `probs` output — in which case label resolution falls back\n *   to whatever the fusion recorded.\n */\nfunction classifierClasses(session: OrtSession): number | null {\n    const index = session.outputNames.indexOf(OUTPUT_PROBS);\n    if (index < 0) return null;\n    const shape = session.outputShapes[index];\n    if (shape === undefined || shape.length === 0) return null;\n    return shape[shape.length - 1] ?? null;\n}\n\n/**\n * Build a class id → name record from an ordered label list.\n *\n * @param labels Labels indexed by class id.\n * @returns The equivalent record.\n */\nfunction indexNames(labels: readonly string[]): Readonly<Record<number, string>> {\n    const names: Record<number, string> = {};\n    for (let i = 0; i < labels.length; i++) names[i] = labels[i] as string;\n    return names;\n}\n\n/**\n * Fetch an output tensor by name.\n *\n * @param outputs The run's results.\n * @param name The output's name in the pipeline contract.\n * @returns The tensor.\n * @throws {@link FusionError} when the graph does not carry that output, which\n *   means the file is not a pipeline this version can drive.\n */\nfunction output(outputs: Record<string, ort.Tensor>, name: string): ort.Tensor {\n    const tensor = outputs[name];\n    if (tensor === undefined) {\n        throw new FusionError(`The fused pipeline is missing its '${name}' output.`);\n    }\n    return tensor;\n}\n\n/**\n * Fetch a float output by name.\n *\n * @param outputs The run's results.\n * @param name The output's name in the pipeline contract.\n * @returns The output's data.\n * @throws {@link FusionError} when the graph does not carry that output.\n */\nfunction floats(outputs: Record<string, ort.Tensor>, name: string): Float32Array {\n    return asFloat32Array(output(outputs, name).data);\n}\n\n/**\n * Fetch an integer output by name, normalizing ORT's 64-bit representation.\n *\n * ONNX Runtime Web returns `int64` tensors as `BigInt64Array`, whose values do\n * not compare or index like numbers. Class ids and detection counts are always\n * small, so widening them to `number` here is lossless and keeps every caller\n * free of `BigInt` handling.\n *\n * @param outputs The run's results.\n * @param name The output's name in the pipeline contract.\n * @returns The output's values as plain numbers.\n * @throws {@link FusionError} when the graph does not carry that output.\n */\nfunction integers(outputs: Record<string, ort.Tensor>, name: string): number[] {\n    const data = output(outputs, name).data as BigInt64Array | Int32Array | Float32Array;\n    const values: number[] = [];\n    for (let i = 0; i < data.length; i++) values.push(Number(data[i]));\n    return values;\n}\n\n/**\n * Convert an image to the `[1, 3, H, W]` float32 tensor the graph expects.\n *\n * @param image The image to convert.\n * @returns The batched CHW tensor, scaled to `[0, 1]`.\n */\nfunction tensorOf(image: RGBImage): ort.Tensor {\n    const chw = toCHW(toFloat32(image), image.width, image.height, 3);\n    return toFloat32Tensor(chw, [1, 3, image.height, image.width]);\n}\n\n/**\n * Assemble one detection, filling the Ultralytics-style aliases.\n *\n * @param classId Detector class index.\n * @param className Detector class name.\n * @param confidence Detection score.\n * @param bbox Box in original-image coordinates.\n * @param croppedImage The region the box covers.\n * @param classification What the classification stage said about that region.\n * @returns The detection object.\n */\nfunction detection(\n    classId: number,\n    className: string,\n    confidence: number,\n    bbox: BoundingBox,\n    croppedImage: RGBImage,\n    classification: ClassificationResult,\n    mask: Mask | null,\n): DetectionResult {\n    return {\n        classId,\n        className,\n        confidence,\n        bbox,\n        cls: classId,\n        name: className,\n        conf: confidence,\n        box: bbox,\n        croppedImage,\n        classification,\n        mask,\n    };\n}\n\n/**\n * Resample a crop-space mask onto the cropped image's own pixel grid.\n *\n * Exported for the test suite rather than for callers: it is not re-exported\n * from the package root, and a mask reaches user code already resampled, on\n * {@link DetectionResult.mask}.\n *\n * The graph computes masks at the resolution the segmenter was exported at,\n * which is rarely the size of the box in the original image. Resampling here is\n * what lets {@link DetectionResult.mask} carry the contract `Segmenter` already\n * produces, so a caller that handles one handles the other.\n *\n * Nearest-neighbour on purpose: the values are already thresholded, and\n * interpolating between them would invent edge pixels that are neither.\n *\n * @param masks The graph's `masks` output.\n * @param row Which detection row to read.\n * @param cropWidth Width of one mask in the graph's output.\n * @param cropHeight Height of one mask in the graph's output.\n * @param target The cropped image the mask must line up with.\n * @returns The mask on the crop's grid, or `null` when the box has no area.\n */\nexport function maskToBox(\n    masks: Float32Array,\n    row: number,\n    cropWidth: number,\n    cropHeight: number,\n    target: RGBImage,\n): Mask | null {\n    const { width, height } = target;\n    if (width < 1 || height < 1) return null;\n    const stride = cropWidth * cropHeight;\n    const source = masks.subarray(row * stride, (row + 1) * stride);\n    const data = new Uint8Array(width * height);\n    for (let y = 0; y < height; y++) {\n        const sourceRow = Math.min(Math.floor((y * cropHeight) / height), cropHeight - 1);\n        for (let x = 0; x < width; x++) {\n            const sourceColumn = Math.min(Math.floor((x * cropWidth) / width), cropWidth - 1);\n            data[y * width + x] =\n                (source[sourceRow * cropWidth + sourceColumn] ?? 0) > 0.5 ? 255 : 0;\n        }\n    }\n    return new Mask(data, width, height);\n}\n\n/**\n * Assemble the bulk-array `Boxes` view from per-instance detections.\n *\n * @param detections The surviving detections.\n * @param origShape `[height, width]` of the original image.\n * @returns The bulk view, empty when nothing survived.\n */\nfunction bulkBoxes(\n    detections: readonly DetectionResult[],\n    origShape: readonly [number, number],\n): Boxes {\n    const xyxy = new Float32Array(detections.length * 4);\n    const cls = new Int32Array(detections.length);\n    const conf = new Float32Array(detections.length);\n    detections.forEach((entry, index) => {\n        const [x1, y1, x2, y2] = entry.bbox.xyxy;\n        xyxy[index * 4] = x1;\n        xyxy[index * 4 + 1] = y1;\n        xyxy[index * 4 + 2] = x2;\n        xyxy[index * 4 + 3] = y2;\n        cls[index] = entry.classId;\n        conf[index] = entry.confidence;\n    });\n    return new Boxes(xyxy, cls, conf, origShape);\n}\n"],"mappings":"waAwGA,IAAa,EAAb,MAAa,UAAuB,EAAA,UAAW,CAGtB,MACA,QACA,OACA,kBACA,iBACA,cAPrB,YACI,EACA,EACA,EACA,EACA,EACA,EACA,EACF,CACE,MAAM,CAAO,EAPI,KAAA,MAAA,EACA,KAAA,QAAA,EACA,KAAA,OAAA,EACA,KAAA,kBAAA,EACA,KAAA,iBAAA,EACA,KAAA,cAAA,CAGrB,CAEA,eAAmD,KAanD,MAAM,OAAO,EAAe,EAAkB,CAC1C,GAAM,CAAC,EAAO,GAAU,KAAK,MAAM,UACnC,IAAK,IAAI,EAAI,EAAG,EAAI,EAAM,IAAK,CAC3B,IAAM,EAAoC,EACrC,EAAA,aAAc,EAAA,gBAAgB,EAAA,eAAe,EAAO,CAAM,EAAG,CAC1D,EACA,EACA,EACA,CACJ,CAAC,CACL,EACI,KAAK,MAAM,mBACX,EAAM,EAAA,cAAgB,EAAA,gBAAgB,EAAA,eAAe,EAAO,CAAM,EAAG,CACjE,EACA,EACA,EACA,CACJ,CAAC,EACD,EAAM,EAAA,aAAe,EAAA,gBAAgB,IAAI,aAAa,CAAC,CAAC,CAAC,EAAG,CAAC,CAAC,CAAC,EAC/D,EAAM,EAAA,WAAa,EAAA,gBAAgB,IAAI,aAAa,CAAC,EAAG,CAAC,CAAC,EAAG,CAAC,CAAC,CAAC,GAEpE,MAAM,KAAK,SAAS,IAAI,CAAK,CACjC,CACJ,CASA,IAAY,WAA+B,CACvC,GAAI,KAAK,iBAAmB,KAAM,CAC9B,GAAM,CAAC,EAAO,GAAU,KAAK,MAAM,UACnC,KAAK,eAAiB,IAAI,EAAA,kBAAkB,EAAO,CAAM,CAC7D,CACA,OAAO,KAAK,cAChB,CAWA,aAAa,OACT,EACA,EAAiC,CAAC,EACX,CACvB,IAAM,EAAU,MAAM,EAAA,WAAW,OAAO,EAAO,CAAO,EAChD,EAAO,EAAA,eAAe,EAAQ,QAAQ,EAC5C,GAAI,IAAS,KACT,MAAM,IAAI,EAAA,YACN,mNAGJ,EAGJ,IAAM,EAAS,EAAA,cAAc,EAAQ,QAAU,EAAK,eAAiB,MAAM,EACrE,EAAmB,EAAA,cAAc,EAAQ,kBAAoB,EAAK,gBAAiB,CACrF,WAAY,EAAkB,CAAO,GAAK,IAAA,EAC9C,CAAC,EACD,OAAO,IAAI,EACP,EACA,EACA,EACA,EAAW,CAAM,EACjB,EACA,EAAW,CAAgB,EAC3B,EAAQ,cAAgB,EAC5B,CACJ,CAGA,IAAI,MAAmB,CACnB,OAAO,KAAK,KAChB,CAGA,IAAI,WAAuC,CACvC,OAAO,KAAK,MAAM,SACtB,CAGA,IAAI,QAA4B,CAC5B,OAAO,KAAK,OAChB,CAGA,IAAI,OAA0C,CAC1C,OAAO,KAAK,MAChB,CAGA,IAAI,kBAAsC,CACtC,OAAO,KAAK,iBAChB,CAGA,IAAI,iBAAoD,CACpD,OAAO,KAAK,gBAChB,CAQA,MAAM,KACF,EACA,EAAwC,CAAC,EACT,CAChC,OAAO,KAAK,QAAQ,EAAO,CAAO,CACtC,CAUA,MAAM,QACF,EACA,EAAwC,CAAC,EACT,CAChC,IAAM,EAAQ,IAAI,EAAA,WACZ,EAAO,OAAO,GAAU,SAAW,EAAQ,KAC3C,CAAE,MAAO,EAAU,UAAW,MAAM,EAAA,gBAAgB,CAAK,EAC/D,EAAM,MAAM,MAAM,EAClB,GAAM,CAAE,QAAO,QAAO,UAAS,SAAQ,UAAW,KAAK,YAAY,EAAU,CAAM,EACnF,EAAM,MAAM,YAAY,EACxB,IAAI,EACJ,GAAI,CACA,EAAU,MAAM,KAAK,SAAS,IAAI,CAAK,CAC3C,QAAU,CACN,KAAK,UAAU,QAAQ,CAAE,QAAO,CAAC,CACrC,CACA,EAAM,MAAM,WAAW,EAEvB,IAAM,EAAc,EAAO,EAAS,EAAA,YAAY,EAC1C,EAAQ,EAAO,EAAS,EAAA,YAAY,EACpC,EAAS,EAAO,EAAS,EAAA,aAAa,EACtC,EAAU,EAAS,EAAS,EAAA,cAAc,EAC1C,EAAQ,EAAA,eAAe,EAAY,IAAI,EACvC,EAAW,EAAS,EAAA,gBAA8B,CAAC,CAAC,IAAM,EAC1D,EAAQ,KAAK,MAAM,SAAW,EAAO,EAAS,EAAA,YAAY,EAAI,KAC9D,CAAC,EAAW,GAAc,KAAK,MAAM,SACrC,EAAO,KAAK,IAAI,EAAU,KAAK,MAAM,EAAM,OAAS,CAAC,CAAC,EACtD,EAAa,EAAY,KAAK,EAAY,KAAK,OAAS,IAAM,EAE9D,EAAU,EAAQ,UAAY,IAAA,GAAY,KAAO,IAAI,IAAI,EAAQ,OAAO,EACxE,EAAQ,EAAQ,eAAiB,EACjC,EAAgC,CAAC,EACvC,IAAK,IAAI,EAAM,EAAG,EAAM,EAAM,IAAO,CACjC,IAAM,EAAU,EAAQ,IAAQ,EAC1B,EAAa,EAAO,IAAQ,EAClC,GAAI,EAAa,GAAU,IAAY,MAAQ,CAAC,EAAQ,IAAI,CAAO,EAAI,SAEvE,IAAM,EAAO,KAAK,YAAY,EAAO,EAAK,CAAE,QAAO,UAAS,SAAQ,UAAS,CAAC,EACxE,EAAU,EAAA,UAAU,EAAU,CAAI,EACxC,EAAW,KACP,EACI,EACA,KAAK,OAAO,IAAY,SAAS,IACjC,EACA,EACA,EACA,KAAK,UACD,EAAM,SAAS,EAAM,GAAa,EAAM,GAAK,CAAU,EACvD,EACA,EAAQ,IACZ,EACA,IAAU,KAAO,KAAO,EAAU,EAAO,EAAK,EAAW,EAAY,CAAO,CAChF,CACJ,CACJ,CAEA,EAAA,kBAAkB,EAAW,OAAQ,CACjC,aAAc,EAAQ,cAAgB,KAAK,cAC3C,cAAe,KAAK,IAAI,EAAO,KAAK,MAAM,aAAa,EACvD,QAAS,EAAQ,QACjB,MACJ,CAAC,EAED,IAAM,EAAuC,CAAC,EAAS,OAAQ,EAAS,KAAK,EAE7E,OADA,EAAM,MAAM,aAAa,EAClB,CACH,IAAI,EAAA,sBACA,EAAU,EAAY,CAAS,EAC/B,EACA,KAAK,OACL,KAAK,iBACL,EACA,EACA,EACA,EAAM,MAAM,CAChB,CACJ,CACJ,CAqBA,YACI,EACA,EAOF,CACE,GAAM,CAAC,EAAO,GAAU,KAAK,MAAM,UAC7B,EAAQ,KAAK,UAAU,IAAI,EAAO,CAAM,EACxC,EAAoC,EACrC,EAAA,aAAc,EAAA,gBAAgB,EAAM,KAAM,CAAC,EAAG,EAAG,EAAQ,CAAK,CAAC,CACpE,EAUA,OARI,KAAK,MAAM,mBACX,EAAM,EAAA,cAAgB,EAAS,CAAK,EACpC,EAAM,EAAA,aAAe,EAAA,gBAAgB,IAAI,aAAa,CAAC,EAAM,KAAK,CAAC,EAAG,CAAC,CAAC,CAAC,EACzE,EAAM,EAAA,WAAa,EAAA,gBACf,IAAI,aAAa,CAAC,EAAM,QAAS,EAAM,MAAM,CAAC,EAC9C,CAAC,CAAC,CACN,GAEG,CACH,QACA,MAAO,EAAM,MACb,QAAS,EAAM,QACf,OAAQ,EAAM,OACd,OAAQ,EAAM,MAClB,CACJ,CAQA,YACI,EACA,EACA,EAMW,CACX,GAAM,CAAE,QAAO,UAAS,SAAQ,YAAa,EACvC,EAAM,GAA2B,EAAM,EAAM,EAAI,IAAW,EAC5D,EAAU,GAA0B,KAAK,IAAI,KAAK,IAAI,EAAO,CAAC,EAAG,EAAS,KAAK,EAC/E,EAAU,GAA0B,KAAK,IAAI,KAAK,IAAI,EAAO,CAAC,EAAG,EAAS,MAAM,EACtF,OAAO,IAAI,EAAA,YACP,GAAQ,EAAG,CAAC,EAAI,GAAW,CAAK,EAChC,GAAQ,EAAG,CAAC,EAAI,GAAU,CAAK,EAC/B,GAAQ,EAAG,CAAC,EAAI,GAAW,CAAK,EAChC,GAAQ,EAAG,CAAC,EAAI,GAAU,CAAK,CACnC,CACJ,CAUA,UACI,EACA,EACA,EACoB,CACpB,IAAM,EAAS,KAAK,MAAM,aAAe,EAAA,QAAQ,CAAG,EAAI,EAClD,CAAE,UAAS,UAAW,EAAA,KAAK,EAAQ,GAAK,IAAI,EAC5C,EAAoC,CAAC,EAC3C,IAAK,IAAI,EAAI,EAAG,EAAI,EAAQ,OAAQ,IAAK,CACrC,IAAM,EAAU,EAAQ,IAAM,EACxB,EAAc,EAAO,IAAM,EAC3B,EAAY,KAAK,kBAAkB,IAAY,SAAS,IAC9D,EAAc,KAAK,CACf,UACA,YACA,cACA,IAAK,EACL,KAAM,EACN,KAAM,CACV,CAAC,CACL,CACA,IAAM,EAAM,EAAc,IAAM,CAC5B,QAAS,EACT,UAAW,UACX,YAAa,EACb,IAAK,EACL,KAAM,UACN,KAAM,CACV,EACA,MAAO,CACH,QAAS,EAAI,QACb,UAAW,EAAI,UACf,WAAY,EAAI,YAChB,IAAK,EAAI,QACT,KAAM,EAAI,UACV,KAAM,EAAI,YACV,QACA,eACJ,CACJ,CACJ,EAUA,SAAS,EAAkB,EAAoC,CAC3D,IAAM,EAAQ,EAAQ,YAAY,QAAQ,EAAA,YAAY,EACtD,GAAI,EAAQ,EAAG,OAAO,KACtB,IAAM,EAAQ,EAAQ,aAAa,GAEnC,OADI,IAAU,IAAA,IAAa,EAAM,SAAW,EAAU,KAC/C,EAAM,EAAM,OAAS,IAAM,IACtC,CAQA,SAAS,EAAW,EAA6D,CAC7E,IAAM,EAAgC,CAAC,EACvC,IAAK,IAAI,EAAI,EAAG,EAAI,EAAO,OAAQ,IAAK,EAAM,GAAK,EAAO,GAC1D,OAAO,CACX,CAWA,SAAS,EAAO,EAAqC,EAA0B,CAC3E,IAAM,EAAS,EAAQ,GACvB,GAAI,IAAW,IAAA,GACX,MAAM,IAAI,EAAA,YAAY,sCAAsC,EAAK,UAAU,EAE/E,OAAO,CACX,CAUA,SAAS,EAAO,EAAqC,EAA4B,CAC7E,OAAO,EAAA,eAAe,EAAO,EAAS,CAAI,CAAC,CAAC,IAAI,CACpD,CAeA,SAAS,EAAS,EAAqC,EAAwB,CAC3E,IAAM,EAAO,EAAO,EAAS,CAAI,CAAC,CAAC,KAC7B,EAAmB,CAAC,EAC1B,IAAK,IAAI,EAAI,EAAG,EAAI,EAAK,OAAQ,IAAK,EAAO,KAAK,OAAO,EAAK,EAAE,CAAC,EACjE,OAAO,CACX,CAQA,SAAS,EAAS,EAA6B,CAC3C,IAAM,EAAM,EAAA,MAAM,EAAA,UAAU,CAAK,EAAG,EAAM,MAAO,EAAM,OAAQ,CAAC,EAChE,OAAO,EAAA,gBAAgB,EAAK,CAAC,EAAG,EAAG,EAAM,OAAQ,EAAM,KAAK,CAAC,CACjE,CAaA,SAAS,EACL,EACA,EACA,EACA,EACA,EACA,EACA,EACe,CACf,MAAO,CACH,UACA,YACA,aACA,OACA,IAAK,EACL,KAAM,EACN,KAAM,EACN,IAAK,EACL,eACA,iBACA,MACJ,CACJ,CAwBA,SAAgB,EACZ,EACA,EACA,EACA,EACA,EACW,CACX,GAAM,CAAE,QAAO,UAAW,EAC1B,GAAI,EAAQ,GAAK,EAAS,EAAG,OAAO,KACpC,IAAM,EAAS,EAAY,EACrB,EAAS,EAAM,SAAS,EAAM,GAAS,EAAM,GAAK,CAAM,EACxD,EAAO,IAAI,WAAW,EAAQ,CAAM,EAC1C,IAAK,IAAI,EAAI,EAAG,EAAI,EAAQ,IAAK,CAC7B,IAAM,EAAY,KAAK,IAAI,KAAK,MAAO,EAAI,EAAc,CAAM,EAAG,EAAa,CAAC,EAChF,IAAK,IAAI,EAAI,EAAG,EAAI,EAAO,IAAK,CAC5B,IAAM,EAAe,KAAK,IAAI,KAAK,MAAO,EAAI,EAAa,CAAK,EAAG,EAAY,CAAC,EAChF,EAAK,EAAI,EAAQ,IACZ,EAAO,EAAY,EAAY,IAAiB,GAAK,GAAM,IAAM,CAC1E,CACJ,CACA,OAAO,IAAI,EAAA,KAAK,EAAM,EAAO,CAAM,CACvC,CASA,SAAS,EACL,EACA,EACK,CACL,IAAM,EAAO,IAAI,aAAa,EAAW,OAAS,CAAC,EAC7C,EAAM,IAAI,WAAW,EAAW,MAAM,EACtC,EAAO,IAAI,aAAa,EAAW,MAAM,EAU/C,OATA,EAAW,SAAS,EAAO,IAAU,CACjC,GAAM,CAAC,EAAI,EAAI,EAAI,GAAM,EAAM,KAAK,KACpC,EAAK,EAAQ,GAAK,EAClB,EAAK,EAAQ,EAAI,GAAK,EACtB,EAAK,EAAQ,EAAI,GAAK,EACtB,EAAK,EAAQ,EAAI,GAAK,EACtB,EAAI,GAAS,EAAM,QACnB,EAAK,GAAS,EAAM,UACxB,CAAC,EACM,IAAI,EAAA,MAAM,EAAM,EAAK,EAAM,CAAS,CAC/C"}