VOC2012 dataset: Absence of aug.txt file while training

ZiqinZhou66 / ZegCLIP

Official implement of CVPR2023 ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation

MIT License

195 stars 17 forks source link

Open aliman80 opened 1 year ago

aliman80 commented 1 year ago

Hi, While trianing on VOC2012 dataset there is no aug.txt file present in the dataset. Can you provide to help train the model better.

Regards

antoineedy commented 2 months ago

Hi! Here is a link that helped me with the dataset augmentation if you are still struggling :)