Semantic Segmentation with SegFormer
19 repos
Semantic segmentation models built on the SegFormer architecture, a transformer-based approach for pixel-level image classification. The cluster centers on fine-tuned SegFormer variants (B0, B2, B5) trained on standard datasets like ADE20K and Cityscapes, demonstrating different model scales and resolution configurations. Repositories here provide pre-trained checkpoints, inference implementations, and documentation for applying efficient vision transformers to dense prediction tasks in computer vision.
Python · 3
transformers ·14,706
image-segmentation ·14,706
pytorch ·14,653
segformer ·12,630
image-processing ·11,744
segmentation ·11,736
computer-vision ·11,732
deeplab-v3-plus ·11,732
models ·11,732
fpn ·11,732
