Image Understanding with Deep Neural Networks

This project involves developing computer vision models to solve challenging problems, such as multi-modal image retrieval, medical imaging or biometrics (especially face understanding and verification), using the new AI models. One focus will be on self-supervised techniques.
Interested students must have taken CS412-Machine Learning course or the equivalent Some deep learning knowledge via online courses etc is also expected. Computer vision course is a plus. Code will be developed in Google Colab or Kaggle.