lukemelas / deep-spectral-segmentation

[CVPR 2022] Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization

Repository from Github https://github.comlukemelas/deep-spectral-segmentation

lukemelas/deep-spectral-segmentation Issues

Demo on Hugging Face Spaces Not Working
Updated 6 months ago
non-executable training code for semantic segmentation
Updated 2 years ago2
error in voc.py when running train.py for semantic segmentation
Updated 2 years ago3
sup_loss and con_loss are not defined in line 237 of train.py. What is purpose of it?
Updated 2 years ago
CRF Segmentations are entirely black
Closed 2 years ago1
Class name of evaluation results
Closed 2 years ago1
semantic-segmentation produces black images
Closed 3 years ago7
Question about extract_eigs
Closed 3 years ago8
Object Localization Problem
Closed 3 years ago2
Size of segmap
Closed 3 years ago3
CRF up-sampled images are blank
Closed 3 years ago5
pymatting module missing in requirements.txt
Closed 3 years ago1
typo in semantic segmentation example - "dino_vitb16" instead of "dino_vits16"
Closed 3 years ago1
Image matting
Updated 3 years ago1
Twice conversion to float for `image_lr`
Closed 3 years ago1
Discussion: can the method differentiate the background classes?
Closed 3 years ago2
Object-localization on VOC2007
Updated 3 years ago
Eigenvector effect
Closed 3 years ago3
problem of dino on semantic-segmentation
Closed 3 years ago3
Details regarding baselines (Saliency-DINO-ViT-B and MaskContrast-DINO-ViT-B)
Closed 3 years ago2