Metadata-Version: 2.1
Name: crazy-vision
Version: 0.1.0
Summary: A video decoding and model inference package using YOLO, OpenCV, and PyTorch
Home-page: https://github.com/yourusername/crazy-vision
Author: Your Name
Author-email: your.email@example.com
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: av
Requires-Dist: opencv-python
Requires-Dist: torch
Requires-Dist: numpy
Requires-Dist: requests
Requires-Dist: ultralytics

# crazy-vision

A video decoding and model inference package that utilizes YOLO, OpenCV, and PyTorch. The package provides classes for GPU/CPU video decoding, frame processing (including resizing and letterboxing), and model inference using YOLO with convenient API wrappers.

## Features

- **Video Decoding:** Supports GPU decoding (NVDEC/CUVID) with a CPU fallback.
- **Frame Processing:** Resize and pad frames while preserving aspect ratio.
- **Model Inference:** Object tracking with YOLO, including batch processing.
- **Easy Integration:** Designed to be modular and extendable.

## Installation

```bash
pip install crazy-vision
