This work evaluates ''Think-with-Videos'' (TwV) agents that resample frames and perform multi-turn reasoning at fixed budgets, and introduces Video* Bench, a benchmark with 2,212 real-world videos and ten tasks spanning content, semantic, and spatio-temporal perception.
Rui-Zhe Chen, Zhi-Ting Fan, Yang Shi et al.· Proceedings of the 34th ACM...· 1 citation
Interactive agents need fine-grained long-video understanding. Current Video-LLMs remain brittle when relevant evidence is subtle or sparse. Fixed frame budgets force rich streams into a few sampled frames and often discard decisive cues. We introduce Video* Bench, a benchmark with 2,212 real-world videos and ten tasks...
Rui-Zhe Chen, Zhi-Ting Fan, Yang Shi et al.· Proceedings of the 34th ACM...· 1 citation
Experiments on video temporal grounding and highlight detection benchmarks show that TEMPURA substantially improves strong base VLMs across model families and scales, confirming that combining event-level reasoning with fine-grained temporal segmentation is an effective recipe for video temporal understanding.