Investigating Video Reasoning Capability Of Large Language Models With Tropes In Movies
Su Hung-ting, Chao Chun-tong, Hsu Ya-ching, Lin Xudong, Niu Yulei, Lee Hung-yi, Hsu Winston H.. Arxiv 2024
[Paper]
[Code]
Large Language Models (LLMs) have demonstrated effectiveness not only in language tasks but also in video reasoning. This paper introduces a novel dataset, Tropes in Movies (TiM), designed as a testbed for exploring two critical yet previously overlooked video reasoning skills: (1) Abstract Perception: understanding and tokenizing abstract concepts in videos, and (2) Long-range Compositional Reasoning: planning and integrating intermediate reasoning steps for understanding long-range videos with numerous frames. Utilizing tropes from movie storytelling, TiM evaluates the reasoning capabilities of state-of-the-art LLM-based approaches. Our experiments show that current methods, including Captioner-Reasoner, Large Multimodal Model Instruction Fine-tuning, and Visual Programming, only marginally outperform a random baseline when tackling the challenges of Abstract Perception and Long-range Compositional Reasoning. To address these deficiencies, we propose Face-Enhanced Viper of Role Interactions (FEVoRI) and Context Query Reduction (ConQueR), which enhance Visual Programming by fostering role interaction awareness and progressively refining movie contexts and trope queries during reasoning processes, significantly improving performance by 15 F1 points. However, this performance still lags behind human levels (40 vs. 65 F1). Additionally, we introduce a new protocol to evaluate the necessity of Abstract Perception and Long-range Compositional Reasoning for task resolution. This is done by analyzing the code generated through Visual Programming using an Abstract Syntax Tree (AST), thereby confirming the increased complexity of TiM. The dataset and code are available at: https://ander1119.github.io/TiM
Similar Work