argus-cv 1.3.0__tar.gz → 1.5.0__tar.gz

{argus_cv-1.3.0 → argus_cv-1.5.0}/CHANGELOG.md

@@ -2,6 +2,22 @@
 
 <!-- version list -->
 
+## v1.5.0 (2026-01-28)
+
+### Features
+
+- Support Roboflow COCO format and improve YOLO classification detection
+  ([`902a59f`](https://github.com/pirnerjonas/argus/commit/902a59fbb506aa586d8677312d21ae240585b511))
+
+
+## v1.4.0 (2026-01-26)
+
+### Features
+
+- **view**: Show polygon vertices and hide bbox for segmentation
+  ([`03b34f7`](https://github.com/pirnerjonas/argus/commit/03b34f78fdafefd35a8b2594bf1296007716d6c9))
+
+
 ## v1.3.0 (2026-01-24)
 
 ### Features

{argus_cv-1.3.0 → argus_cv-1.5.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: argus-cv
-Version: 1.3.0
+Version: 1.5.0
 Summary: CLI tool for working with vision AI datasets
 Requires-Python: >=3.10
 Requires-Dist: numpy>=1.24.0
@@ -12,7 +12,7 @@ Description-Content-Type: text/markdown
 
 # argus-cv
 
-Vision AI dataset toolkit for working with YOLO and COCO datasets.
+Vision AI dataset toolkit for working with YOLO, COCO, and semantic mask datasets.
 
 **[Documentation](https://pirnerjonas.github.io/argus/)**
 
@@ -22,6 +22,13 @@ Vision AI dataset toolkit for working with YOLO and COCO datasets.
 uvx argus-cv
 ```
 
+## Highlights
+
+- Detects YOLO, COCO, and folder-based semantic mask datasets.
+- Reports per-class stats (pixel coverage for mask datasets).
+- Interactive viewer with bounding boxes, polygons, or mask overlays.
+- Split unsplit YOLO/COCO datasets into train/val/test.
+
 ## Usage
 
 ```bash

{argus_cv-1.3.0 → argus_cv-1.5.0}/README.md

@@ -1,6 +1,6 @@
 # argus-cv
 
-Vision AI dataset toolkit for working with YOLO and COCO datasets.
+Vision AI dataset toolkit for working with YOLO, COCO, and semantic mask datasets.
 
 **[Documentation](https://pirnerjonas.github.io/argus/)**
 
@@ -10,6 +10,13 @@ Vision AI dataset toolkit for working with YOLO and COCO datasets.
 uvx argus-cv
 ```
 
+## Highlights
+
+- Detects YOLO, COCO, and folder-based semantic mask datasets.
+- Reports per-class stats (pixel coverage for mask datasets).
+- Interactive viewer with bounding boxes, polygons, or mask overlays.
+- Split unsplit YOLO/COCO datasets into train/val/test.
+
 ## Usage
 
 ```bash

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/getting-started/installation.md

@@ -6,7 +6,7 @@ workflow.
 ## Quick install with uv
 
 ```bash
-uvx argus
+uvx argus-cv
 ```
 
 `uvx` runs the package in an isolated environment and keeps it up to date.
@@ -14,19 +14,19 @@ uvx argus
 Verify it works:
 
 ```bash
-argus --help
+argus-cv --help
 ```
 
 ## pipx
 
 ```bash
-pipx install argus
+pipx install argus-cv
 ```
 
 Verify it works:
 
 ```bash
-argus --help
+argus-cv --help
 ```
 
 ## From source
@@ -40,7 +40,7 @@ pip install -e .
 Verify it works:
 
 ```bash
-argus --help
+argus-cv --help
 ```
 
 ## Requirements

argus_cv-1.5.0/docs/guides/datasets.md

@@ -0,0 +1,142 @@
+# Dataset formats
+
+Argus supports YOLO, COCO, and folder-based semantic mask datasets. Detection
+and segmentation are handled out of the box.
+
+## YOLO
+
+Argus looks for a YAML config file with a `names` key. It uses that file to
+extract class names and verify the dataset layout.
+
+Typical structure:
+
+```text
+dataset/
+├── data.yaml
+├── images/
+│   ├── train/
+│   ├── val/
+│   └── test/
+└── labels/
+    ├── train/
+    ├── val/
+    └── test/
+```
+
+Unsplit YOLO datasets are also supported:
+
+```text
+dataset/
+├── data.yaml
+├── images/
+└── labels/
+```
+
+Argus infers the task type by scanning a few label files:
+
+- 5 values per line: detection
+- more than 5 values per line: segmentation polygons
+
+## COCO
+
+Argus looks for COCO annotation JSON files in `annotations/` or at the dataset
+root.
+
+Typical structure:
+
+```text
+dataset/
+├── annotations/
+│   ├── instances_train.json
+│   ├── instances_val.json
+│   └── instances_test.json
+└── images/
+    ├── train/
+    ├── val/
+    └── test/
+```
+
+If your annotation filenames include `train`, `val`, or `test`, Argus will treat
+those as splits. Otherwise it defaults to `train`.
+
+### Roboflow COCO
+
+Argus also supports the Roboflow variant of COCO format, where annotations live
+inside split directories:
+
+```text
+dataset/
+├── train/
+│   ├── _annotations.coco.json
+│   └── *.jpg
+├── valid/
+│   ├── _annotations.coco.json
+│   └── *.jpg
+└── test/
+    ├── _annotations.coco.json
+    └── *.jpg
+```
+
+Splits are detected from directory names (`train`, `valid`/`val`, `test`).
+
+## Mask (semantic segmentation)
+
+Mask datasets are simple image + mask folders. Argus detects a few common
+patterns:
+
+- `images/` + `masks/`
+- `img/` + `gt/`
+- `leftImg8bit/` + `gtFine/` (Cityscapes-style)
+
+Split-aware layout:
+
+```text
+dataset/
+├── images/
+│   ├── train/
+│   └── val/
+├── masks/
+│   ├── train/
+│   └── val/
+└── classes.yaml  # Optional for grayscale, required for RGB palette masks
+```
+
+Unsplit layout:
+
+```text
+dataset/
+├── images/
+├── masks/
+└── classes.yaml
+```
+
+### Mask encoding
+
+- Grayscale masks: each pixel value is the class ID. Argus will auto-detect
+  class IDs if no `classes.yaml` is provided.
+- RGB palette masks: each class maps to a color. A `classes.yaml` is required.
+- Mask files should be `.png` and match the image stem (e.g., `frame.png`), or
+  use common suffixes like `_mask`, `_gt`, or `_label`.
+
+Example `classes.yaml`:
+
+```yaml
+names:
+  - background
+  - road
+  - sidewalk
+ignore_index: 255
+palette:
+  - id: 0
+    name: background
+  - id: 1
+    name: road
+```
+
+## Detection heuristics
+
+If Argus does not detect your dataset, check the following:
+
+- The dataset root is correct and readable.
+- YOLO: the YAML file includes `names` as a list or dict.
+- COCO: `images`, `annotations`, and `categories` exist in the JSON.

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/guides/listing.md

@@ -11,7 +11,7 @@ argus-cv list --path /datasets
 This prints a table showing:
 
 - Dataset path
-- Format (yolo or coco)
+- Format (yolo, coco, or mask)
 - Task (detection or segmentation)
 - Number of classes
 - Detected splits

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/guides/splitting.md

@@ -8,7 +8,7 @@ Use `argus-cv split` to create train/val/test splits from an unsplit dataset.
 argus-cv split -d /datasets/animals -o /datasets/animals_splits
 ```
 
-By default, Argus uses a 0.8/0.1/0.1 ratio and stratified sampling.
+Argus uses a 0.8/0.1/0.1 ratio and stratified sampling by default.
 
 ## Custom ratio
 
@@ -18,12 +18,6 @@ argus-cv split -d /datasets/animals -o /datasets/animals_splits -r 0.7,0.2,0.1
 
 Ratios can sum to 1.0 or 100.
 
-## Disable stratification
-
-```bash
-argus-cv split -d /datasets/animals -o /datasets/animals_splits --no-stratify
-```
-
 ## Set a seed for determinism
 
 ```bash

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/guides/stats.md

@@ -1,6 +1,8 @@
 # Stats and counts
 
 `argus-cv stats` provides per-class instance counts and image totals by split.
+For mask datasets, it reports pixel coverage and how many images contain each
+class.
 
 ## Example
 
@@ -25,3 +27,4 @@ If Argus prints "No annotations found", check:
 
 - YOLO: `labels/` exists and matches `images/`.
 - COCO: annotation JSON files are valid and contain `annotations`.
+- Mask: masks are `.png` files and match the image file names.

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/guides/viewer.md

@@ -1,6 +1,7 @@
 # Visual inspection
 
-The viewer overlays boxes and masks for quick spot checks.
+The viewer overlays boxes and masks for quick spot checks. For mask datasets,
+it blends the segmentation mask over the image.
 
 ## Launching the viewer
 
@@ -14,6 +15,12 @@ argus-cv view -d /datasets/retail
 argus-cv view -d /datasets/retail --split val
 ```
 
+### Adjust mask opacity
+
+```bash
+argus-cv view -d /datasets/roads --opacity 0.3
+```
+
 ## Controls
 
 - Right arrow or `N`: next image
@@ -22,6 +29,7 @@ argus-cv view -d /datasets/retail --split val
 - Drag: pan while zoomed
 - `R`: reset zoom
 - `Q` or `Esc`: quit
+- `T`: toggle annotations or mask overlay
 
 ## Notes
 

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/index.md

@@ -3,8 +3,9 @@
     <p class="hero__eyebrow">Argus</p>
     <h1>Vision AI dataset work, without the friction.</h1>
     <p>
-      Argus is a focused CLI for YOLO and COCO datasets. List datasets, inspect
-      class balance, view annotations, and split cleanly for training.
+      Argus is a focused CLI for YOLO, COCO, and semantic mask datasets. List
+      datasets, inspect class balance, view annotations, and split cleanly for
+      training.
     </p>
     <div class="hero__actions">
       <a class="md-button md-button--primary" href="getting-started/quickstart/">Get started</a>
@@ -30,7 +31,7 @@ argus-cv view -d /data/animals --split val
 <div class="grid cards">
   <div class="card">
     <h3>Format-aware</h3>
-    <p>Detects YOLO and COCO by structure and metadata, not guesses.</p>
+    <p>Detects YOLO, COCO, and mask datasets by structure and metadata.</p>
   </div>
   <div class="card">
     <h3>Readable statistics</h3>
@@ -38,7 +39,7 @@ argus-cv view -d /data/animals --split val
   </div>
   <div class="card">
     <h3>Annotation viewer</h3>
-    <p>Browse images with boxes and masks overlayed. Pan and zoom included.</p>
+    <p>Browse images with boxes, polygons, and mask overlays. Pan and zoom included.</p>
   </div>
   <div class="card">
     <h3>Clean splits</h3>

{argus_cv-1.3.0 → argus_cv-1.5.0}/docs/reference/cli.md

@@ -3,7 +3,7 @@
 ## Global
 
 ```bash
-argus --help
+argus-cv --help
 ```
 
 Argus uses subcommands: `list`, `stats`, `view`, and `split`.
@@ -45,6 +45,7 @@ Options:
 
 - `--dataset-path`, `-d`: dataset root path
 - `--split`, `-s`: split to view (train, val, test)
+- `--opacity`, `-o`: mask overlay opacity (mask datasets only)
 
 ## split
 
@@ -54,7 +55,6 @@ Create train/val/test splits from an unsplit dataset.
 argus-cv split --dataset-path /datasets/animals \
   --output-path /datasets/animals_splits \
   --ratio 0.8,0.1,0.1 \
-  --stratify \
   --seed 42
 ```
 
@@ -63,5 +63,4 @@ Options:
 - `--dataset-path`, `-d`: dataset root path
 - `--output-path`, `-o`: output directory (default: "splits" inside dataset path)
 - `--ratio`, `-r`: train/val/test ratio (default: 0.8,0.1,0.1)
-- `--stratify/--no-stratify`: enable or disable stratified splitting (default: enabled)
 - `--seed`: random seed (default: 42)

{argus_cv-1.3.0 → argus_cv-1.5.0}/mkdocs.yml

@@ -1,5 +1,5 @@
 site_name: Argus
-site_description: Vision AI dataset toolkit for YOLO and COCO datasets.
+site_description: Vision AI dataset toolkit for YOLO, COCO, and mask datasets.
 site_author: Argus
 site_url: https://pirnerjonas.github.io/argus/
 repo_name: pirnerjonas/argus

{argus_cv-1.3.0 → argus_cv-1.5.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "argus-cv"
-version = "1.3.0"
+version = "1.5.0"
 description = "CLI tool for working with vision AI datasets"
 readme = "README.md"
 requires-python = ">=3.10"

{argus_cv-1.3.0 → argus_cv-1.5.0}/src/argus/__init__.py

@@ -1,3 +1,3 @@
 """Argus - Vision AI dataset toolkit."""
 
-__version__ = "1.3.0"
+__version__ = "1.5.0"

{argus_cv-1.3.0 → argus_cv-1.5.0}/src/argus/cli.py

@@ -8,11 +8,18 @@ import cv2
 import numpy as np
 import typer
 from rich.console import Console
-from rich.progress import Progress, SpinnerColumn, TextColumn
+from rich.progress import (
+    BarColumn,
+    Progress,
+    SpinnerColumn,
+    TaskProgressColumn,
+    TextColumn,
+)
 from rich.table import Table
 
 from argus.core import COCODataset, Dataset, MaskDataset, YOLODataset
 from argus.core.base import DatasetFormat, TaskType
+from argus.core.convert import convert_mask_to_yolo_seg
 from argus.core.split import (
     is_coco_unsplit,
     parse_ratio,
@@ -534,13 +541,6 @@ def split_dataset(
             help="Train/val/test ratio (e.g. 0.8,0.1,0.1).",
         ),
     ] = "0.8,0.1,0.1",
-    stratify: Annotated[
-        bool,
-        typer.Option(
-            "--stratify/--no-stratify",
-            help="Stratify by class distribution when splitting.",
-        ),
-    ] = True,
     seed: Annotated[
         int,
         typer.Option(
@@ -593,9 +593,7 @@ def split_dataset(
         ) as progress:
             progress.add_task("Creating YOLO splits...", total=None)
             try:
-                counts = split_yolo_dataset(
-                    dataset, output_path, ratios, stratify, seed
-                )
+                counts = split_yolo_dataset(dataset, output_path, ratios, True, seed)
             except ValueError as exc:
                 console.print(f"[red]Error: {exc}[/red]")
                 raise typer.Exit(1) from exc
@@ -628,7 +626,7 @@ def split_dataset(
                     annotation_file,
                     output_path,
                     ratios,
-                    stratify,
+                    True,
                     seed,
                 )
             except ValueError as exc:
@@ -641,6 +639,148 @@ def split_dataset(
     )
 
 
+@app.command(name="convert")
+def convert_dataset(
+    input_path: Annotated[
+        Path,
+        typer.Option(
+            "--input-path",
+            "-i",
+            help="Path to the source dataset.",
+        ),
+    ] = Path("."),
+    output_path: Annotated[
+        Path,
+        typer.Option(
+            "--output-path",
+            "-o",
+            help="Output directory for converted dataset.",
+        ),
+    ] = Path("converted"),
+    to_format: Annotated[
+        str,
+        typer.Option(
+            "--to",
+            help="Target format (currently only 'yolo-seg' is supported).",
+        ),
+    ] = "yolo-seg",
+    epsilon_factor: Annotated[
+        float,
+        typer.Option(
+            "--epsilon-factor",
+            "-e",
+            help="Polygon simplification factor (Douglas-Peucker algorithm).",
+            min=0.0,
+            max=1.0,
+        ),
+    ] = 0.005,
+    min_area: Annotated[
+        float,
+        typer.Option(
+            "--min-area",
+            "-a",
+            help="Minimum contour area in pixels to include.",
+            min=0.0,
+        ),
+    ] = 100.0,
+) -> None:
+    """Convert a dataset from one format to another.
+
+    Currently supports converting MaskDataset to YOLO segmentation format.
+
+    Example:
+        uvx argus-cv convert -i /path/to/masks -o /path/to/output --to yolo-seg
+    """
+    # Validate format
+    if to_format != "yolo-seg":
+        console.print(
+            f"[red]Error: Unsupported target format '{to_format}'.[/red]\n"
+            "[yellow]Currently only 'yolo-seg' is supported.[/yellow]"
+        )
+        raise typer.Exit(1)
+
+    # Resolve and validate input path
+    input_path = input_path.resolve()
+    if not input_path.exists():
+        console.print(f"[red]Error: Path does not exist: {input_path}[/red]")
+        raise typer.Exit(1)
+    if not input_path.is_dir():
+        console.print(f"[red]Error: Path is not a directory: {input_path}[/red]")
+        raise typer.Exit(1)
+
+    # Detect source dataset - must be MaskDataset for yolo-seg conversion
+    dataset = MaskDataset.detect(input_path)
+    if not dataset:
+        console.print(
+            f"[red]Error: No MaskDataset found at {input_path}[/red]\n"
+            "[yellow]Ensure the path contains images/ + masks/ directories "
+            "(or equivalent patterns like img/+gt/ or leftImg8bit/+gtFine/).[/yellow]"
+        )
+        raise typer.Exit(1)
+
+    # Resolve output path
+    if not output_path.is_absolute():
+        output_path = input_path.parent / output_path
+    output_path = output_path.resolve()
+
+    # Check if output already exists
+    if output_path.exists() and any(output_path.iterdir()):
+        console.print(
+            f"[red]Error: Output directory already exists and is not empty: "
+            f"{output_path}[/red]"
+        )
+        raise typer.Exit(1)
+
+    # Show conversion info
+    console.print("[cyan]Converting MaskDataset to YOLO segmentation format[/cyan]")
+    console.print(f"  Source: {input_path}")
+    console.print(f"  Output: {output_path}")
+    console.print(f"  Classes: {dataset.num_classes}")
+    splits_str = ", ".join(dataset.splits) if dataset.splits else "unsplit"
+    console.print(f"  Splits: {splits_str}")
+    console.print()
+
+    # Run conversion with progress bar
+    with Progress(
+        SpinnerColumn(),
+        TextColumn("[progress.description]{task.description}"),
+        BarColumn(),
+        TaskProgressColumn(),
+        console=console,
+    ) as progress:
+        task = progress.add_task("Processing images...", total=None)
+
+        def update_progress(current: int, total: int) -> None:
+            progress.update(task, completed=current, total=total)
+
+        try:
+            stats = convert_mask_to_yolo_seg(
+                dataset=dataset,
+                output_path=output_path,
+                epsilon_factor=epsilon_factor,
+                min_area=min_area,
+                progress_callback=update_progress,
+            )
+        except Exception as exc:
+            console.print(f"[red]Error during conversion: {exc}[/red]")
+            raise typer.Exit(1) from exc
+
+    # Show results
+    console.print()
+    console.print("[green]Conversion complete![/green]")
+    console.print(f"  Images processed: {stats['images']}")
+    console.print(f"  Labels created: {stats['labels']}")
+    console.print(f"  Polygons extracted: {stats['polygons']}")
+
+    if stats["skipped"] > 0:
+        skipped = stats["skipped"]
+        console.print(f"  [yellow]Skipped: {skipped} (no mask or empty)[/yellow]")
+    if stats["warnings"] > 0:
+        console.print(f"  [yellow]Warnings: {stats['warnings']}[/yellow]")
+
+    console.print(f"\n[cyan]Output dataset: {output_path}[/cyan]")
+
+
 class _ImageViewer:
     """Interactive image viewer with zoom and pan support."""
 
@@ -1352,9 +1492,12 @@ def _draw_annotations(
             overlay = img.copy()
             cv2.fillPoly(overlay, [pts], color)
             cv2.addWeighted(overlay, 0.3, img, 0.7, 0, img)
+            # Draw small points at polygon vertices
+            for pt in pts:
+                cv2.circle(img, tuple(pt), radius=3, color=color, thickness=-1)
 
-        # Draw bounding box
-        if bbox:
+        # Draw bounding box (only for detection, not segmentation)
+        if bbox and not polygon:
             x, y, w, h = bbox
             x1, y1 = int(x), int(y)
             x2, y2 = int(x + w), int(y + h)

{argus_cv-1.3.0 → argus_cv-1.5.0}/src/argus/core/__init__.py

@@ -2,6 +2,13 @@
 
 from argus.core.base import Dataset
 from argus.core.coco import COCODataset
+from argus.core.convert import (
+    ConversionParams,
+    Polygon,
+    convert_mask_to_yolo_labels,
+    convert_mask_to_yolo_seg,
+    mask_to_polygons,
+)
 from argus.core.mask import ConfigurationError, MaskDataset
 from argus.core.split import split_coco_dataset, split_yolo_dataset
 from argus.core.yolo import YOLODataset
@@ -14,4 +21,9 @@ __all__ = [
     "ConfigurationError",
     "split_coco_dataset",
     "split_yolo_dataset",
+    "ConversionParams",
+    "Polygon",
+    "mask_to_polygons",
+    "convert_mask_to_yolo_labels",
+    "convert_mask_to_yolo_seg",
 ]