Open-Vocabulary Camouflaged Object Segmentation with Cascaded Vision Language Models

Prepare work

1. sam-vit-h-pth: https://dl.fbaipublicfiles.com/segment_anything/sam_vit_h_4b8939.pth

2. alpha-clip-best-model.pth and best-model.pth: link: https://pan.baidu.com/s/1q2XueAeRyFBend4ilGIQ2w?pwd=siy7 password: siy7

Put the 'sam-vit-h-pth' and 'alpha-clip-best-model.pth' in the folder pretrained.

Put the 'best-model.pth' in the folder best_model_pth.

Test

python test_ovsam_maskdecoder1_edge.py