Abstract: Text-based Visual Question Answering (TextVQA) focuses on answering questions about the scene text in images. Most works in this field uses transformer based models to modeling the ...
Abstract: Low-quality pseudo labels pose a significant obstacle in semi-supervised medical image segmentation (SSMIS), impeding consistency learning on unlabeled data. Leveraging vision-language model ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results