rtmlib-ts 0.0.2

package/src/solution/poseDetector.ts

@@ -0,0 +1,892 @@
+/**
+ * PoseDetector - Unified API for person detection and pose estimation
+ * Combines YOLO12 detector with RTMW pose model in a single optimized interface
+ *
+ * @example
+ * ```typescript
+ * // Initialize with default models (from HuggingFace)
+ * const detector = new PoseDetector();
+ * await detector.init();
+ *
+ * // Or with custom models
+ * const detector = new PoseDetector({
+ *   detModel: 'models/yolov12n.onnx',
+ *   poseModel: 'models/rtmlib/end2end.onnx',
+ * });
+ * await detector.init();
+ *
+ * // From canvas
+ * const results = await detector.detectFromCanvas(canvas);
+ *
+ * // From video element
+ * const results = await detector.detectFromVideo(videoElement);
+ *
+ * // From raw image data
+ * const results = await detector.detect(imageData, width, height);
+ * ```
+ */
+
+import * as ort from 'onnxruntime-web';
+import { BBox, Detection } from '../types/index';
+import { getCachedModel, isModelCached } from '../core/modelCache';
+
+// Configure ONNX Runtime Web
+ort.env.wasm.wasmPaths = 'https://cdn.jsdelivr.net/npm/onnxruntime-web@1.23.0/dist/';
+ort.env.wasm.simd = true;
+ort.env.wasm.proxy = false;
+
+/**
+ * Configuration options for PoseDetector
+ */
+export interface PoseDetectorConfig {
+  /** Path to YOLO12 detection model (optional - uses default from HuggingFace if not specified) */
+  detModel?: string;
+  /** Path to RTMW pose estimation model (optional - uses default from HuggingFace if not specified) */
+  poseModel?: string;
+  /** Detection input size (default: [416, 416]) */
+  detInputSize?: [number, number];
+  /** Pose input size (default: [384, 288]) */
+  poseInputSize?: [number, number];
+  /** Detection confidence threshold (default: 0.5) */
+  detConfidence?: number;
+  /** NMS IoU threshold (default: 0.45) */
+  nmsThreshold?: number;
+  /** Pose keypoint confidence threshold (default: 0.3) */
+  poseConfidence?: number;
+  /** Execution backend (default: 'wasm') */
+  backend?: 'wasm' | 'webgpu';
+  /** Enable model caching (default: true) */
+  cache?: boolean;
+}
+
+/**
+ * Detected person with bounding box and keypoints
+ */
+export interface Person {
+  /** Bounding box coordinates */
+  bbox: {
+    x1: number;
+    y1: number;
+    x2: number;
+    y2: number;
+    confidence: number;
+  };
+  /** 17 COCO keypoints coordinates */
+  keypoints: Keypoint[];
+  /** Keypoint scores (0-1) */
+  scores: number[];
+}
+
+/**
+ * Single keypoint with coordinates and visibility
+ */
+export interface Keypoint {
+  x: number;
+  y: number;
+  score: number;
+  visible: boolean;
+  name: string;
+}
+
+/**
+ * Detection statistics
+ */
+export interface PoseStats {
+  /** Number of detected people */
+  personCount: number;
+  /** Detection inference time (ms) */
+  detTime: number;
+  /** Pose estimation time (ms) */
+  poseTime: number;
+  /** Total processing time (ms) */
+  totalTime: number;
+}
+
+/**
+ * COCO17 keypoint names
+ */
+const KEYPOINT_NAMES = [
+  'nose',
+  'left_eye',
+  'right_eye',
+  'left_ear',
+  'right_ear',
+  'left_shoulder',
+  'right_shoulder',
+  'left_elbow',
+  'right_elbow',
+  'left_wrist',
+  'right_wrist',
+  'left_hip',
+  'right_hip',
+  'left_knee',
+  'right_knee',
+  'left_ankle',
+  'right_ankle',
+];
+
+/**
+ * Default configuration
+ */
+const DEFAULT_CONFIG: Required<PoseDetectorConfig> = {
+  detModel: 'https://huggingface.co/demon2233/rtmlib-ts/resolve/main/yolo/yolov12n.onnx',
+  poseModel: 'https://huggingface.co/demon2233/rtmlib-ts/resolve/main/rtmpose/end2end.onnx',
+  detInputSize: [416, 416],  // Faster detection
+  poseInputSize: [384, 288],  // Required by model
+  detConfidence: 0.5,
+  nmsThreshold: 0.45,
+  poseConfidence: 0.3,
+  backend: 'webgpu',  // Default to WebGPU for better performance
+  cache: true,
+};
+
+export class PoseDetector {
+  private config: Required<PoseDetectorConfig>;
+  private detSession: ort.InferenceSession | null = null;
+  private poseSession: ort.InferenceSession | null = null;
+  private initialized = false;
+
+  // Pre-allocated buffers for maximum performance
+  private canvas: HTMLCanvasElement | null = null;
+  private ctx: CanvasRenderingContext2D | null = null;
+  private poseCanvas: HTMLCanvasElement | null = null;
+  private poseCtx: CanvasRenderingContext2D | null = null;
+  private poseTensorBuffer: Float32Array | null = null;
+  private detInputSize: [number, number] = [416, 416];
+  private poseInputSize: [number, number] = [384, 288];
+
+  constructor(config: PoseDetectorConfig) {
+    this.config = { ...DEFAULT_CONFIG, ...config };
+  }
+
+  /**
+   * Initialize both detection and pose models with pre-allocated resources
+   */
+  async init(): Promise<void> {
+    if (this.initialized) return;
+
+    try {
+      // Load detection model
+      console.log(`[PoseDetector] Loading detection model from: ${this.config.detModel}`);
+      let detBuffer: ArrayBuffer;
+      
+      if (this.config.cache) {
+        const detCached = await isModelCached(this.config.detModel);
+        console.log(`[PoseDetector] Det model cache ${detCached ? 'hit' : 'miss'}`);
+        detBuffer = await getCachedModel(this.config.detModel);
+      } else {
+        const detResponse = await fetch(this.config.detModel);
+        if (!detResponse.ok) {
+          throw new Error(`Failed to fetch det model: HTTP ${detResponse.status}`);
+        }
+        detBuffer = await detResponse.arrayBuffer();
+      }
+      
+      this.detSession = await ort.InferenceSession.create(detBuffer, {
+        executionProviders: [this.config.backend],
+        graphOptimizationLevel: 'all',
+      });
+      console.log(`[PoseDetector] Detection model loaded, size: ${(detBuffer.byteLength / 1024 / 1024).toFixed(2)} MB`);
+
+      // Load pose model
+      console.log(`[PoseDetector] Loading pose model from: ${this.config.poseModel}`);
+      let poseBuffer: ArrayBuffer;
+      
+      if (this.config.cache) {
+        const poseCached = await isModelCached(this.config.poseModel);
+        console.log(`[PoseDetector] Pose model cache ${poseCached ? 'hit' : 'miss'}`);
+        poseBuffer = await getCachedModel(this.config.poseModel);
+      } else {
+        const poseResponse = await fetch(this.config.poseModel);
+        if (!poseResponse.ok) {
+          throw new Error(`Failed to fetch pose model: HTTP ${poseResponse.status}`);
+        }
+        poseBuffer = await poseResponse.arrayBuffer();
+      }
+      
+      this.poseSession = await ort.InferenceSession.create(poseBuffer, {
+        executionProviders: [this.config.backend],
+        graphOptimizationLevel: 'all',
+      });
+      console.log(`[PoseDetector] Pose model loaded, size: ${(poseBuffer.byteLength / 1024 / 1024).toFixed(2)} MB`);
+
+      // Pre-allocate all resources
+      const [detW, detH] = this.config.detInputSize;
+      this.detInputSize = [detW, detH];
+
+      const [poseW, poseH] = this.config.poseInputSize;
+      this.poseInputSize = [poseW, poseH];
+
+      // Main canvas for detection
+      this.canvas = document.createElement('canvas');
+      this.canvas.width = detW;
+      this.canvas.height = detH;
+      this.ctx = this.canvas.getContext('2d', {
+        willReadFrequently: true,
+        alpha: false
+      })!;
+
+      // Pose crop canvas (reused for each person)
+      this.poseCanvas = document.createElement('canvas');
+      this.poseCanvas.width = poseW;
+      this.poseCanvas.height = poseH;
+      this.poseCtx = this.poseCanvas.getContext('2d', {
+        willReadFrequently: true,
+        alpha: false
+      })!;
+
+      // Pre-allocate pose tensor buffer
+      this.poseTensorBuffer = new Float32Array(3 * poseW * poseH);
+
+      this.initialized = true;
+      console.log(`[PoseDetector] ✅ Initialized (det:${detW}x${detH}, pose:${poseW}x${poseH})`);
+    } catch (error) {
+      console.error('[PoseDetector] ❌ Initialization failed:', error);
+      throw error;
+    }
+  }
+
+  /**
+   * Detect poses from HTMLCanvasElement
+   * @param canvas - Canvas element containing the image
+   * @returns Array of detected people with keypoints
+   */
+  async detectFromCanvas(canvas: HTMLCanvasElement): Promise<Person[]> {
+    const ctx = canvas.getContext('2d');
+    if (!ctx) {
+      throw new Error('Could not get 2D context from canvas');
+    }
+
+    const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
+    return this.detect(new Uint8Array(imageData.data.buffer), canvas.width, canvas.height);
+  }
+
+  /**
+   * Detect poses from HTMLVideoElement
+   * @param video - Video element to capture frame from
+   * @param targetCanvas - Optional canvas for frame extraction (creates one if not provided)
+   * @returns Array of detected people with keypoints
+   */
+  async detectFromVideo(
+    video: HTMLVideoElement,
+    targetCanvas?: HTMLCanvasElement
+  ): Promise<Person[]> {
+    if (video.readyState < 2) {
+      throw new Error('Video not ready. Ensure video is loaded and playing.');
+    }
+
+    const canvas = targetCanvas || document.createElement('canvas');
+    canvas.width = video.videoWidth;
+    canvas.height = video.videoHeight;
+
+    const ctx = canvas.getContext('2d');
+    if (!ctx) {
+      throw new Error('Could not get 2D context from canvas');
+    }
+
+    ctx.drawImage(video, 0, 0, canvas.width, canvas.height);
+    const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
+
+    return this.detect(new Uint8Array(imageData.data.buffer), canvas.width, canvas.height);
+  }
+
+  /**
+   * Detect poses from HTMLImageElement
+   * @param image - Image element to process
+   * @param targetCanvas - Optional canvas for image extraction (creates one if not provided)
+   * @returns Array of detected people with keypoints
+   */
+  async detectFromImage(
+    image: HTMLImageElement,
+    targetCanvas?: HTMLCanvasElement
+  ): Promise<Person[]> {
+    if (!image.complete || !image.naturalWidth) {
+      throw new Error('Image not loaded. Ensure image is fully loaded.');
+    }
+
+    const canvas = targetCanvas || document.createElement('canvas');
+    canvas.width = image.naturalWidth;
+    canvas.height = image.naturalHeight;
+
+    const ctx = canvas.getContext('2d');
+    if (!ctx) {
+      throw new Error('Could not get 2D context from canvas');
+    }
+
+    ctx.drawImage(image, 0, 0);
+    const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
+
+    return this.detect(new Uint8Array(imageData.data.buffer), canvas.width, canvas.height);
+  }
+
+  /**
+   * Detect poses from ImageBitmap (efficient for blob/file uploads)
+   * @param bitmap - ImageBitmap to process
+   * @param targetCanvas - Optional canvas for bitmap extraction (creates one if not provided)
+   * @returns Array of detected people with keypoints
+   */
+  async detectFromBitmap(
+    bitmap: ImageBitmap,
+    targetCanvas?: HTMLCanvasElement
+  ): Promise<Person[]> {
+    const canvas = targetCanvas || document.createElement('canvas');
+    canvas.width = bitmap.width;
+    canvas.height = bitmap.height;
+
+    const ctx = canvas.getContext('2d');
+    if (!ctx) {
+      throw new Error('Could not get 2D context from canvas');
+    }
+
+    ctx.drawImage(bitmap, 0, 0);
+    const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
+
+    return this.detect(new Uint8Array(imageData.data.buffer), canvas.width, canvas.height);
+  }
+
+  /**
+   * Detect poses from File (for file input uploads)
+   * @param file - File object from input element
+   * @param targetCanvas - Optional canvas for image extraction (creates one if not provided)
+   * @returns Array of detected people with keypoints
+   */
+  async detectFromFile(
+    file: File,
+    targetCanvas?: HTMLCanvasElement
+  ): Promise<Person[]> {
+    return new Promise((resolve, reject) => {
+      const img = new Image();
+      img.onload = async () => {
+        try {
+          const results = await this.detectFromImage(img, targetCanvas);
+          resolve(results);
+        } catch (error) {
+          reject(error);
+        }
+      };
+      img.onerror = () => reject(new Error('Failed to load image from file'));
+      img.src = URL.createObjectURL(file);
+    });
+  }
+
+  /**
+   * Detect poses from Blob (for camera capture or downloads)
+   * @param blob - Blob object to process
+   * @param targetCanvas - Optional canvas for image extraction (creates one if not provided)
+   * @returns Array of detected people with keypoints
+   */
+  async detectFromBlob(
+    blob: Blob,
+    targetCanvas?: HTMLCanvasElement
+  ): Promise<Person[]> {
+    const bitmap = await createImageBitmap(blob);
+    const results = await this.detectFromBitmap(bitmap, targetCanvas);
+    bitmap.close();
+    return results;
+  }
+
+  /**
+   * Detect people and estimate poses in a single call
+   * @param imageData - Image data (Uint8Array RGB/RGBA)
+   * @param width - Image width
+   * @param height - Image height
+   * @returns Array of detected people with keypoints
+   */
+  async detect(
+    imageData: Uint8Array,
+    width: number,
+    height: number
+  ): Promise<Person[]> {
+    if (!this.initialized) {
+      await this.init();
+    }
+
+    const startTime = performance.now();
+
+    // Step 1: Detect people
+    const detStart = performance.now();
+    const bboxes = await this.detectPeople(imageData, width, height);
+    const detTime = performance.now() - detStart;
+
+    // Step 2: Estimate poses for each person
+    const poseStart = performance.now();
+    const people: Person[] = [];
+
+    for (const bbox of bboxes) {
+      const keypoints = await this.estimatePose(imageData, width, height, bbox);
+      people.push({
+        bbox: {
+          x1: bbox.x1,
+          y1: bbox.y1,
+          x2: bbox.x2,
+          y2: bbox.y2,
+          confidence: bbox.confidence,
+        },
+        keypoints,
+        scores: keypoints.map(k => k.score),
+      });
+    }
+
+    const poseTime = performance.now() - poseStart;
+    const totalTime = performance.now() - startTime;
+
+    // Attach stats (for debugging)
+    (people as any).stats = {
+      personCount: people.length,
+      detTime: Math.round(detTime),
+      poseTime: Math.round(poseTime),
+      totalTime: Math.round(totalTime),
+    } as PoseStats;
+
+    return people;
+  }
+
+  /**
+   * Get detection and pose statistics from last call
+   */
+  getStats(): PoseStats | null {
+    return null; // Stats attached to results
+  }
+
+  /**
+   * Detect people using YOLO12
+   */
+  private async detectPeople(
+    imageData: Uint8Array,
+    width: number,
+    height: number
+  ): Promise<Array<{ x1: number; y1: number; x2: number; y2: number; confidence: number }>> {
+    const [inputH, inputW] = this.config.detInputSize;
+
+    // Preprocess
+    const { tensor, paddingX, paddingY, scaleX, scaleY } = this.preprocessYOLO(
+      imageData,
+      width,
+      height,
+      [inputW, inputH]
+    );
+
+    // Inference - use dynamic input name
+    const inputTensor = new ort.Tensor('float32', tensor, [1, 3, inputH, inputW]);
+    const inputName = this.detSession!.inputNames[0];  // Dynamic: 'images' or 'pixel_values'
+    
+    const feeds: Record<string, ort.Tensor> = {};
+    feeds[inputName] = inputTensor;
+    
+    const results = await this.detSession!.run(feeds);
+    const output = results[this.detSession!.outputNames[0]];
+
+    // Postprocess
+    return this.postprocessYOLO(
+      output.data as Float32Array,
+      output.dims[1],
+      width,
+      height,
+      paddingX,
+      paddingY,
+      scaleX,
+      scaleY
+    );
+  }
+
+  /**
+   * Estimate pose for a single person
+   */
+  private async estimatePose(
+    imageData: Uint8Array,
+    imgWidth: number,
+    imgHeight: number,
+    bbox: { x1: number; y1: number; x2: number; y2: number; confidence: number }
+  ): Promise<Keypoint[]> {
+    const [inputH, inputW] = this.config.poseInputSize;
+
+    // Preprocess
+    const { tensor, center, scale } = this.preprocessPose(
+      imageData,
+      imgWidth,
+      imgHeight,
+      bbox,
+      [inputW, inputH]
+    );
+
+    // Inference
+    const inputTensor = new ort.Tensor('float32', tensor, [1, 3, inputH, inputW] as number[]);
+    const results = await this.poseSession!.run({ input: inputTensor });
+
+    // Postprocess
+    return this.postprocessPose(
+      results.simcc_x.data as Float32Array,
+      results.simcc_y.data as Float32Array,
+      results.simcc_x.dims as number[],
+      results.simcc_y.dims as number[],
+      center,
+      scale
+    );
+  }
+
+  /**
+   * YOLO preprocessing with letterbox
+   */
+  private preprocessYOLO(
+    imageData: Uint8Array,
+    imgWidth: number,
+    imgHeight: number,
+    inputSize: [number, number]
+  ): {
+    tensor: Float32Array;
+    paddingX: number;
+    paddingY: number;
+    scaleX: number;
+    scaleY: number;
+  } {
+    const [inputW, inputH] = inputSize;
+
+    // Reuse canvas
+    if (!this.canvas || !this.ctx) {
+      this.canvas = document.createElement('canvas');
+      this.ctx = this.canvas.getContext('2d', { willReadFrequently: true })!;
+    }
+
+    this.canvas.width = inputW;
+    this.canvas.height = inputH;
+    const ctx = this.ctx;
+
+    // Black background
+    ctx.fillStyle = '#000000';
+    ctx.fillRect(0, 0, inputW, inputH);
+
+    // Calculate letterbox
+    const aspectRatio = imgWidth / imgHeight;
+    const targetAspectRatio = inputW / inputH;
+
+    let drawWidth: number, drawHeight: number, offsetX: number, offsetY: number;
+
+    if (aspectRatio > targetAspectRatio) {
+      drawWidth = inputW;
+      drawHeight = Math.floor(inputW / aspectRatio);
+      offsetX = 0;
+      offsetY = Math.floor((inputH - drawHeight) / 2);
+    } else {
+      drawHeight = inputH;
+      drawWidth = Math.floor(inputH * aspectRatio);
+      offsetX = Math.floor((inputW - drawWidth) / 2);
+      offsetY = 0;
+    }
+
+    // Create source canvas
+    const srcCanvas = document.createElement('canvas');
+    const srcCtx = srcCanvas.getContext('2d')!;
+    srcCanvas.width = imgWidth;
+    srcCanvas.height = imgHeight;
+
+    const srcImageData = srcCtx.createImageData(imgWidth, imgHeight);
+    srcImageData.data.set(imageData);
+    srcCtx.putImageData(srcImageData, 0, 0);
+
+    // Draw
+    ctx.drawImage(srcCanvas, 0, 0, imgWidth, imgHeight, offsetX, offsetY, drawWidth, drawHeight);
+
+    const paddedData = ctx.getImageData(0, 0, inputW, inputH);
+
+    // Normalize to [0, 1] and convert to CHW
+    const tensor = new Float32Array(inputW * inputH * 3);
+    for (let i = 0; i < paddedData.data.length; i += 4) {
+      const pixelIdx = i / 4;
+      tensor[pixelIdx] = paddedData.data[i] / 255;
+      tensor[pixelIdx + inputW * inputH] = paddedData.data[i + 1] / 255;
+      tensor[pixelIdx + 2 * inputW * inputH] = paddedData.data[i + 2] / 255;
+    }
+
+    const scaleX = imgWidth / drawWidth;
+    const scaleY = imgHeight / drawHeight;
+
+    return {
+      tensor,
+      paddingX: offsetX,
+      paddingY: offsetY,
+      scaleX,
+      scaleY,
+    };
+  }
+
+  /**
+   * YOLO postprocessing with NMS
+   */
+  private postprocessYOLO(
+    output: Float32Array,
+    numDetections: number,
+    imgWidth: number,
+    imgHeight: number,
+    paddingX: number,
+    paddingY: number,
+    scaleX: number,
+    scaleY: number
+  ): Array<{ x1: number; y1: number; x2: number; y2: number; confidence: number }> {
+    const detections: Array<{ x1: number; y1: number; x2: number; y2: number; confidence: number }> = [];
+
+    for (let i = 0; i < numDetections; i++) {
+      const idx = i * 6;
+      const x1 = output[idx];
+      const y1 = output[idx + 1];
+      const x2 = output[idx + 2];
+      const y2 = output[idx + 3];
+      const confidence = output[idx + 4];
+      const classId = Math.round(output[idx + 5]);
+
+      if (confidence < this.config.detConfidence || classId !== 0) continue;
+
+      // Transform coordinates
+      const tx1 = (x1 - paddingX) * scaleX;
+      const ty1 = (y1 - paddingY) * scaleY;
+      const tx2 = (x2 - paddingX) * scaleX;
+      const ty2 = (y2 - paddingY) * scaleY;
+
+      detections.push({
+        x1: Math.max(0, tx1),
+        y1: Math.max(0, ty1),
+        x2: Math.min(imgWidth, tx2),
+        y2: Math.min(imgHeight, ty2),
+        confidence,
+      });
+    }
+
+    // NMS
+    return this.applyNMS(detections, this.config.nmsThreshold);
+  }
+
+  /**
+   * Pose preprocessing with affine crop
+   */
+  private preprocessPose(
+    imageData: Uint8Array,
+    imgWidth: number,
+    imgHeight: number,
+    bbox: { x1: number; y1: number; x2: number; y2: number; confidence: number },
+    inputSize: [number, number]
+  ): { tensor: Float32Array; center: [number, number]; scale: [number, number] } {
+    const [inputW, inputH] = inputSize;
+
+    const bboxWidth = bbox.x2 - bbox.x1;
+    const bboxHeight = bbox.y2 - bbox.y1;
+
+    const center: [number, number] = [
+      bbox.x1 + bboxWidth / 2,
+      bbox.y1 + bboxHeight / 2,
+    ];
+
+    // Aspect ratio preservation
+    const bboxAspectRatio = bboxWidth / bboxHeight;
+    const modelAspectRatio = inputW / inputH;
+
+    let scaleW: number, scaleH: number;
+    if (bboxAspectRatio > modelAspectRatio) {
+      scaleW = bboxWidth * 1.25;
+      scaleH = scaleW / modelAspectRatio;
+    } else {
+      scaleH = bboxHeight * 1.25;
+      scaleW = scaleH * modelAspectRatio;
+    }
+
+    const scale: [number, number] = [scaleW, scaleH];
+
+    // Reuse pre-allocated pose canvas
+    if (!this.poseCanvas || !this.poseCtx) {
+      this.poseCanvas = document.createElement('canvas');
+      this.poseCanvas.width = inputW;
+      this.poseCanvas.height = inputH;
+      this.poseCtx = this.poseCanvas.getContext('2d', {
+        willReadFrequently: true,
+        alpha: false
+      })!;
+      this.poseTensorBuffer = new Float32Array(3 * inputW * inputH);
+    }
+
+    const ctx = this.poseCtx;
+
+    // Fast clear
+    ctx.clearRect(0, 0, inputW, inputH);
+
+    // Create source
+    const srcCanvas = document.createElement('canvas');
+    const srcCtx = srcCanvas.getContext('2d')!;
+    srcCanvas.width = imgWidth;
+    srcCanvas.height = imgHeight;
+
+    const srcImageData = srcCtx.createImageData(imgWidth, imgHeight);
+    srcImageData.data.set(imageData);
+    srcCtx.putImageData(srcImageData, 0, 0);
+
+    // Crop and scale
+    const srcX = center[0] - scaleW / 2;
+    const srcY = center[1] - scaleH / 2;
+    ctx.drawImage(srcCanvas, srcX, srcY, scaleW, scaleH, 0, 0, inputW, inputH);
+
+    const croppedData = ctx.getImageData(0, 0, inputW, inputH);
+
+    // Optimized normalization with precomputed constants
+    const tensor = this.poseTensorBuffer!;
+    const len = croppedData.data.length;
+    const planeSize = inputW * inputH;
+    
+    // Precompute normalization constants
+    const mean0 = 123.675, mean1 = 116.28, mean2 = 103.53;
+    const stdInv0 = 1 / 58.395, stdInv1 = 1 / 57.12, stdInv2 = 1 / 57.375;
+    
+    // Unrolled loop (4 pixels at once)
+    for (let i = 0; i < len; i += 16) {
+      const p1 = i / 4, p2 = p1 + 1, p3 = p1 + 2, p4 = p1 + 3;
+      
+      // R channel
+      tensor[p1] = (croppedData.data[i] - mean0) * stdInv0;
+      tensor[p2] = (croppedData.data[i + 4] - mean0) * stdInv0;
+      tensor[p3] = (croppedData.data[i + 8] - mean0) * stdInv0;
+      tensor[p4] = (croppedData.data[i + 12] - mean0) * stdInv0;
+      
+      // G channel
+      tensor[p1 + planeSize] = (croppedData.data[i + 1] - mean1) * stdInv1;
+      tensor[p2 + planeSize] = (croppedData.data[i + 5] - mean1) * stdInv1;
+      tensor[p3 + planeSize] = (croppedData.data[i + 9] - mean1) * stdInv1;
+      tensor[p4 + planeSize] = (croppedData.data[i + 13] - mean1) * stdInv1;
+      
+      // B channel
+      tensor[p1 + planeSize * 2] = (croppedData.data[i + 2] - mean2) * stdInv2;
+      tensor[p2 + planeSize * 2] = (croppedData.data[i + 6] - mean2) * stdInv2;
+      tensor[p3 + planeSize * 2] = (croppedData.data[i + 10] - mean2) * stdInv2;
+      tensor[p4 + planeSize * 2] = (croppedData.data[i + 14] - mean2) * stdInv2;
+    }
+
+    return { tensor, center, scale };
+  }
+
+  /**
+   * Pose postprocessing with SimCC decoding
+   */
+  private postprocessPose(
+    simccX: Float32Array,
+    simccY: Float32Array,
+    shapeX: number[],
+    shapeY: number[],
+    center: [number, number],
+    scale: [number, number]
+  ): Keypoint[] {
+    const numKeypoints = shapeX[1];
+    const wx = shapeX[2];
+    const wy = shapeY[2];
+
+    const keypoints: Keypoint[] = [];
+
+    for (let k = 0; k < numKeypoints; k++) {
+      // Argmax X
+      let maxX = -Infinity;
+      let argmaxX = 0;
+      for (let i = 0; i < wx; i++) {
+        const val = simccX[k * wx + i];
+        if (val > maxX) {
+          maxX = val;
+          argmaxX = i;
+        }
+      }
+
+      // Argmax Y
+      let maxY = -Infinity;
+      let argmaxY = 0;
+      for (let i = 0; i < wy; i++) {
+        const val = simccY[k * wy + i];
+        if (val > maxY) {
+          maxY = val;
+          argmaxY = i;
+        }
+      }
+
+      const score = 0.5 * (maxX + maxY);
+      const visible = score > this.config.poseConfidence;
+
+      // Transform to original coordinates
+      const normX = argmaxX / wx;
+      const normY = argmaxY / wy;
+
+      const x = (normX - 0.5) * scale[0] + center[0];
+      const y = (normY - 0.5) * scale[1] + center[1];
+
+      keypoints.push({
+        x,
+        y,
+        score,
+        visible,
+        name: KEYPOINT_NAMES[k] || `keypoint_${k}`,
+      });
+    }
+
+    return keypoints;
+  }
+
+  /**
+   * Non-Maximum Suppression
+   */
+  private applyNMS(
+    detections: Array<{ x1: number; y1: number; x2: number; y2: number; confidence: number }>,
+    iouThreshold: number
+  ): typeof detections {
+    if (detections.length === 0) return [];
+
+    detections.sort((a, b) => b.confidence - a.confidence);
+
+    const selected: typeof detections = [];
+    const used = new Set<number>();
+
+    for (let i = 0; i < detections.length; i++) {
+      if (used.has(i)) continue;
+
+      selected.push(detections[i]);
+      used.add(i);
+
+      for (let j = i + 1; j < detections.length; j++) {
+        if (used.has(j)) continue;
+
+        const iou = this.calculateIoU(detections[i], detections[j]);
+        if (iou > iouThreshold) {
+          used.add(j);
+        }
+      }
+    }
+
+    return selected;
+  }
+
+  /**
+   * Calculate IoU between two boxes
+   */
+  private calculateIoU(
+    box1: { x1: number; y1: number; x2: number; y2: number },
+    box2: { x1: number; y1: number; x2: number; y2: number }
+  ): number {
+    const x1 = Math.max(box1.x1, box2.x1);
+    const y1 = Math.max(box1.y1, box2.y1);
+    const x2 = Math.min(box1.x2, box2.x2);
+    const y2 = Math.min(box1.y2, box2.y2);
+
+    if (x2 <= x1 || y2 <= y1) return 0;
+
+    const intersection = (x2 - x1) * (y2 - y1);
+    const area1 = (box1.x2 - box1.x1) * (box1.y2 - box1.y1);
+    const area2 = (box2.x2 - box2.x1) * (box2.y2 - box2.y1);
+    const union = area1 + area2 - intersection;
+
+    return intersection / union;
+  }
+
+  /**
+   * Dispose resources
+   */
+  dispose(): void {
+    if (this.detSession) {
+      this.detSession.release();
+      this.detSession = null;
+    }
+    if (this.poseSession) {
+      this.poseSession.release();
+      this.poseSession = null;
+    }
+    this.initialized = false;
+  }
+}