Causal Chaos Paper

Project Brief

Helped to co-author a paper under Dr Paritosh Parmar in IHPC SCC (Social and Cognitive Computing) regarding a new and improved dataset for comprehensive causal action question answering over longer causal chains grounded in dynamic visual scenes.

Accomplishments

• Helped to annotate the dataset and discuss the format of the dataset
• Benchmarked different video question answering (VQA) models against our new dataset
• Evaluated and researched on different reasoning types for the dataset creation
• Accepted as a poster paper (23.29% acceptance) in NeurIps 2024

Technologies and Skills Used:

• Python
• Pytorch
• Dataset Annotation

Gallery

Featured in A*STAR CFAR(Centre for Frontier AI Research) news