sithu31296/semantic-segmentation
A PyTorch library of state-of-the-art semantic segmentation models and vision transformer backbones for scene parsing, human parsing, and medical imaging.
Velocity · 7d
+0.5
★ / day
Trend
→steady
star history
This repository offers easy-to-use implementations of semantic segmentation models in PyTorch, supporting 20+ datasets and 15+ backbone architectures including ResNet variants and vision transformers. It provides training pipelines, model exports to ONNX/TFLite/OpenVINO, and inference utilities. The library covers scene parsing, human parsing, face parsing, and supports custom dataset integration.