search

MendelXu / SAN

Open-vocabulary Semantic Segmentation
https://mendelxu.github.io/SAN/
MIT License
295 stars 27 forks source link

how to get pcontext-full dataset #16

Closed sunwhw closed 1 year ago

sunwhw commented 1 year ago

Thanks for you great work! I met a problem when i prepare the dataset image What value should I set for anno dir? i guess the anno dir should be set "dataset/pcontext-full/VOCdevkit/VOC2010/Annotations" but "xml" is not compatible with 'mat' which assigned in code? what should i do? image

MendelXu commented 1 year ago

You can download it from here. https://cs.stanford.edu/~roozbeh/pascal-context/trainval.tar.gz After you extract the compressed file, you will be able to find a folder trainval. It contains all the .mat format files. image

sunwhw commented 1 year ago

oh! Thanks