[2603.22911] ForestPrune: High-ratio Visual Token Compression for Video Multimodal Large Language Models via Spatial-Temporal Forest Modeling

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.22911 (cs) [Submitted on 24 Mar 2026] Title:ForestPrune: High-ratio Visual Token Compression for Video Multimodal Large Language Models via Spatial-Temporal Forest Modeling Authors:Shaobo Ju, Baiyang Song, Tao Chen, Jiapeng Zhang, Qiong Wu, Chao Chang, HuaiXi Wang, Yiyi Zhou, Rongrong Ji View a PDF of the paper titled ForestPrune: High-ratio Visual Token Compression for Video Multimodal Large Language Models via Spatial-Temporal Forest Modeling, by Shaobo Ju and 8 other authors View PDF HTML (experimental) Abstract:Due to the great saving of computation and memory overhead, token compression has become a research hot-spot for MLLMs and achieved remarkable progress in image-language tasks. However, for the video, existing methods still fall short of high-ratio token compression. We attribute this shortcoming to the insufficient modeling of temporal and continual video content, and propose a novel and training-free token pruning method for video MLLMs, termed ForestPrune, which achieves effective and high-ratio pruning via Spatial-temporal Forest Modeling. In practice, ForestPrune construct token forests across video frames based on the semantic, spatial and temporal constraints, making an overall comprehension of videos. Afterwards, ForestPrune evaluates the importance of token trees and nodes based on tree depth and node roles, thereby obtaining a globally optimal pruning decision. To validate ForestPrun...