PhD student, Chinese University of Hong Kong (Shenzhen)
1 paper at NeurIPS 2025
This work investigates whether MLLMs can be used as unified evaluator for AI-generated videos and proposes a benchmark to assess this ability.