Machine learning for video and image understanding

Description

Recent years has seen significant improvement in the ability for machines to infer the content of images and videos. These improvements have been driven from the increased availability of annotated training data and faster hardware for training and evaluating models coupled with advances in deep learning models.

This is a broad project that explores numerous topics in video and image understanding including human pose estimation, inferring human-object interaction, understanding human activities, and visual question answering (VQA). The topic can be tailored to your background and interests.

 

Updated:  1 November 2018/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing