Gaze based Video Summarization
Created scripts for the paper Gaze-enabled Egocentric Video Summarization via Constrained Submodular Maximization from scratch. Tested out the results in videoes taken from CMU using AR glasses
Topic Covered:
- Basics of CLIP Models and Multimodal ML.
- Backbone - MRCNN.
- Optimization Techniques - Submodular Optimization, Mutual Information.
- Clustering Algorithm - KMeans, Greedy, Temporal Star clustering.
Programming Language: Python, Pytorch