Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/bjornmelin/ml-vision-lab
👁️ Production-grade computer vision implementations. Real-world applications in image processing, object detection, and video analytics with GPU acceleration. 📸
https://github.com/bjornmelin/ml-vision-lab
computer-vision cuda deep-learning image-processing object-detection opencv pytorch video-analytics
Last synced: 19 days ago
JSON representation
👁️ Production-grade computer vision implementations. Real-world applications in image processing, object detection, and video analytics with GPU acceleration. 📸
- Host: GitHub
- URL: https://github.com/bjornmelin/ml-vision-lab
- Owner: BjornMelin
- License: mit
- Created: 2025-01-24T02:48:22.000Z (20 days ago)
- Default Branch: main
- Last Pushed: 2025-01-24T14:37:49.000Z (19 days ago)
- Last Synced: 2025-01-24T15:30:07.100Z (19 days ago)
- Topics: computer-vision, cuda, deep-learning, image-processing, object-detection, opencv, pytorch, video-analytics
- Homepage:
- Size: 2.93 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# ML Vision Lab 👁️
[](https://www.python.org/downloads/)
[](https://pytorch.org/)
[](https://opencv.org/)
[](https://developer.nvidia.com/cuda-toolkit)
[](LICENSE)
> Production-grade computer vision implementations. Advanced applications in image processing, object detection, and video analytics with GPU acceleration.
[Features](#features) • [Installation](#installation) • [Quick Start](#quick-start) • [Documentation](#documentation) • [Contributing](#contributing)
## 📑 Table of Contents
- [Features](#features)
- [Project Structure](#project-structure)
- [Prerequisites](#prerequisites)
- [Installation](#installation)
- [Quick Start](#quick-start)
- [Documentation](#documentation)
- [Models](#models)
- [Pipeline Optimization](#pipeline-optimization)
- [Benchmarks](#benchmarks)
- [Contributing](#contributing)
- [Versioning](#versioning)
- [Authors](#authors)
- [Citation](#citation)
- [License](#license)
- [Acknowledgments](#acknowledgments)
## ✨ Features
- Real-time object detection and tracking
- Image segmentation pipelines
- Video analytics systems
- GPU-accelerated processing
- Production deployment examples
## 📁 Project Structure
```mermaid
graph TD
A[ml-vision-lab] --> B[models]
A --> C[pipelines]
A --> D[utils]
A --> E[deployment]
B --> F[detection]
B --> G[segmentation]
B --> H[tracking]
C --> I[preprocessing]
C --> J[augmentation]
D --> K[visualization]
D --> L[evaluation]
E --> M[serving]
E --> N[optimization]
```
Click to expand full directory structure
```plaintext
ml-vision-lab/
├── models/ # Model implementations
│ ├── detection/ # Object detection models
│ ├── segmentation/ # Segmentation models
│ └── tracking/ # Tracking algorithms
├── pipelines/ # Processing pipelines
│ ├── preprocessing/# Data preprocessing
│ └── augmentation/ # Data augmentation
├── utils/ # Utility functions
├── deployment/ # Deployment tools
├── tests/ # Unit tests
└── README.md # Documentation
```
## 🔧 Prerequisites
- Python 3.8+
- CUDA 11.8+
- OpenCV 4.8+
- PyTorch 2.2+
- NVIDIA GPU (Compute Capability 6.0+)
## 📦 Installation
```bash
# Clone repository
git clone https://github.com/BjornMelin/ml-vision-lab.git
cd ml-vision-lab
# Create environment
python -m venv venv
source venv/bin/activate
# Install dependencies
pip install -r requirements.txt
```
## 🚀 Quick Start
```python
from vision_lab import detection, tracking
# Initialize detector
detector = detection.YOLODetector(
model='yolov8',
confidence=0.5
)
# Setup video pipeline
tracker = tracking.DeepSORT(
max_age=30,
n_init=3
)
# Process video stream
results = detector.process_video(
source='video.mp4',
tracker=tracker
)
```
## 📚 Documentation
### Models
| Model | Task | Performance | Speed (FPS) |
|-------|------|-------------|-------------|
| YOLOv8 | Detection | mAP: 52.3 | 120 |
| Mask R-CNN | Segmentation | mAP: 47.8 | 45 |
| DeepSORT | Tracking | MOTA: 76.5 | 80 |
### Pipeline Optimization
- GPU-accelerated preprocessing
- Batch processing optimization
- Memory-efficient inference
- TensorRT integration
### Benchmarks
Performance on standard datasets:
| Task | Dataset | Model | GPU | FPS | Accuracy |
|------|---------|-------|-----|-----|----------|
| Detection | COCO | YOLOv8 | A100 | 120 | mAP: 52.3 |
| Segmentation | COCO | Mask R-CNN | V100 | 45 | mAP: 47.8 |
| Tracking | MOT17 | DeepSORT | 3090 | 80 | MOTA: 76.5 |
## 🤝 Contributing
- [Contributing Guidelines](CONTRIBUTING.md)
- [Code of Conduct](CODE_OF_CONDUCT.md)
- [Development Guide](DEVELOPMENT.md)
## 📌 Versioning
We use [SemVer](http://semver.org/) for versioning. For available versions, see the [tags on this repository](https://github.com/BjornMelin/ml-vision-lab/tags).
## ✍️ Authors
**Bjorn Melin**
- GitHub: [@BjornMelin](https://github.com/BjornMelin)
- LinkedIn: [Bjorn Melin](https://linkedin.com/in/bjorn-melin)
## 📝 Citation
```bibtex
@misc{melin2024mlvisionlab,
author = {Melin, Bjorn},
title = {ML Vision Lab: Production Computer Vision Implementations},
year = {2024},
publisher = {GitHub},
url = {https://github.com/BjornMelin/ml-vision-lab}
}
```
## 📄 License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
## 🙏 Acknowledgments
- OpenCV community
- YOLO authors and contributors
- Deep SORT implementation team
---
Made with 👁️ and ❤️ by Bjorn Melin