Metadata-Version: 2.4
Name: ort-vision-sdk
Version: 0.2.0
Summary: High-level Python SDK for computer vision inference with ONNX Runtime.
Project-URL: Homepage, https://github.com/mauriciobenjamin700/ort-vision-sdk
Project-URL: Repository, https://github.com/mauriciobenjamin700/ort-vision-sdk
Project-URL: Issues, https://github.com/mauriciobenjamin700/ort-vision-sdk/issues
Project-URL: Changelog, https://github.com/mauriciobenjamin700/ort-vision-sdk/blob/main/CHANGELOG.md
Author-email: Mauricio Benjamin <mauriciobenjamin700@gmail.com>
License-Expression: MIT
License-File: LICENSE
Keywords: computer-vision,inference,onnx,onnxruntime,sdk
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Typing :: Typed
Requires-Python: >=3.10
Requires-Dist: numpy>=1.24.0
Requires-Dist: onnxruntime>=1.17.0
Requires-Dist: pillow>=10.0.0
Provides-Extra: dev
Requires-Dist: build>=1.2.0; extra == 'dev'
Requires-Dist: mypy>=1.10.0; extra == 'dev'
Requires-Dist: pytest-cov>=5.0.0; extra == 'dev'
Requires-Dist: pytest>=8.0.0; extra == 'dev'
Requires-Dist: ruff>=0.5.0; extra == 'dev'
Requires-Dist: twine>=5.0.0; extra == 'dev'
Provides-Extra: gpu
Requires-Dist: onnxruntime-gpu>=1.17.0; extra == 'gpu'
Provides-Extra: opencv
Requires-Dist: opencv-python>=4.8.0; extra == 'opencv'
Description-Content-Type: text/markdown

# ort-vision-sdk

High-level Python SDK for computer vision inference on top of [ONNX Runtime](https://onnxruntime.ai/).

Wraps the low-level `InferenceSession` API with task-oriented classes (`Classifier`, `Detector`, ...) that handle preprocessing, execution provider selection, and postprocessing — so you go from an image to a typed result in one call.

## Installation

```bash
pip install ort-vision-sdk          # CPU only
pip install ort-vision-sdk[gpu]     # CUDA
pip install ort-vision-sdk[opencv]  # adds cv2 image backend
```

## Quick start

```python
from ort_vision_sdk import Classifier

clf = Classifier("resnet50.onnx", labels="imagenet")
result = clf.predict("dog.jpg")

print(result.class_name, result.confidence)
print(result.probabilities[:5])  # top-5 ClassProbability tuples
```

```python
from ort_vision_sdk import Detector

det = Detector("yolov8n.onnx", labels="coco")
detections = det.predict("street.jpg")

for d in detections:
    print(d.class_name, d.confidence, d.bbox.as_xyxy())
    # d.cropped_image is a np.ndarray (HWC, RGB, uint8)
```

## Status

Alpha — API may change. See [`pyproject.toml`](pyproject.toml) for supported Python and dependency versions.
