lxtGH
commented
3 months ago Thanks!
1, We disagree with you. The parts are contributed by the original authors.
2, seems like there are conflicts. Moreover, we welcome the in-submission works with project page and Github link.
Hi, first thank you for your great project! Maybe I am wrong but I think Primitive Generation and Semantic-related Alignment for Universal Zero-Shot Segmentation and ZegFormer: Decoupling Zero-Shot Semantic Segmentation should not belong to Open-Vocabulary Segmentation but Zero-shot Segmentation. If I am wrong, please let me know, and look forward to your reply.
If possible, please consider to add the paper CLIP Is Also a Good Teacher: A New Learning Framework for Inductive Zero-shot Semantic Segmentation and Generalizable Semantic Vision Query Generation for Zero-shot Panoptic and Semantic Segmentation. These two papers can do open-vocabulary segmentations even under the settings of zero-shot segmentation.