foundation models, fine-grained image recognition, vision-language, contrastive learning, multimodal
No public activity