Abstract: The aim of this project is to learn to generalize to unseen classes during training which is an interesting challenge in the CV. Students will learn how to work with multi-modal data (text, image) to train a zero/few-shot model.

Contact: Ali Armin