PeterL1n/RobustVideoMatting
Robust Video Matting in PyTorch, TensorFlow, TensorFlow.js, ONNX, CoreML!
Employs a recurrent neural network architecture with temporal memory to process video frames sequentially rather than independently, enabling real-time matting at 4K 76FPS and HD 104FPS. Offers dual backbone options (MobileNetv3 for efficiency, ResNet50 for accuracy) and supports deployment across PyTorch, TensorFlow, TensorFlow.js, ONNX, CoreML, and TorchScript with quantization variants for mobile optimization.
9,264 stars. No commits in the last 6 months.
Stars
9,264
Forks
1,192
Language
Python
License
GPL-3.0
Category
Last pushed
Apr 02, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/PeterL1n/RobustVideoMatting"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
patrikhuber/eos
A lightweight 3D Morphable Face Model library in modern C++
Uason-Chen/CTR-GCN
[ICCV2021] Official code for "Channel-wise Topology Refinement Graph Convolution for...
PeterL1n/BackgroundMattingV2
Real-Time High-Resolution Background Matting
Zielon/MICA
MICA - Towards Metrical Reconstruction of Human Faces [ECCV2022]
LukasHedegaard/co3d
Official source code for "Continual 3D Convolutional Neural Networks for Real-time Processing of...