This paper presents a multi-agent perception-action exploration alliance, dubbed A4VL, for efficient long-video reasoning. A4VL operates in a multi-round perception-action exploration loop with a selection of VLM agents. In each round, the team of agents performs video question-answer (VideoQA) via perception exploration followed by action exploration. During perception exploration, each agent learns to extract query-specific perception clue(s) from a few sampled frames and performs clue-based alignment to find the video block(s) that are most relevant to the query-specific event. During action exploration, A4VL performs video reasoning in three steps: (1) each agent produces its initial answer with rational, (2) all agents collaboratively scores one another through cross-reviews and relevance ranking, and (3) based on whether a satisfactory consensus is reached, the decision is made either to start a new round of perception-action deliberation by pruning (e.g., filtering out the lowest performing agent) and re-staging (e.g., new-clue and matching block based perception-action exploration), or to conclude by producing its final answer. The integration of the multi-agent alliance through multi-round perception-action exploration, coupled with event-driven partitioning and cue-guided block alignment, enables A4VL to effectively scale to real world long videos while preserving high quality video reasoning. Evaluation Results on five popular VideoQA benchmarks show that A4VL outperforms 18 existing representative VLMs and 11 recent methods optimized for long-video reasoning, while achieving significantly lower inference latency. Our code is released at https://github.com/git-disl/A4VL.

翻译：本文提出一个名为A4VL的多智能体感知-行动探索联盟，用于高效的长视频推理。A4VL通过多轮感知-行动探索循环运作，并选取一组视觉语言模型（VLM）智能体。在每一轮中，智能体团队通过感知探索随后进行行动探索来执行视频问答任务。在感知探索阶段，每个智能体学会从少量采样帧中提取查询特定的感知线索，并基于这些线索进行对齐，以找到与查询特定事件最相关的视频块。在行动探索阶段，A4VL通过三个步骤进行视频推理：（1）每个智能体生成带有论证的初步答案；（2）所有智能体通过交叉评审和相关性排序进行相互评分；（3）根据是否达成令人满意的共识，决定是开启新一轮感知-行动推理（例如，通过剪枝淘汰表现最差的智能体并重组，如基于新线索和匹配块的感知-行动探索），还是输出最终答案并结束推理。通过多轮感知-行动探索整合多智能体联盟，结合事件驱动划分与线索引导的块对齐，A4VL能够有效扩展至真实长视频场景，同时保持高质量的视频推理。在五个主流视频问答基准上的评估结果表明，A4VL的性能优于18个现有的代表性视觉语言模型和11个近期针对长视频推理优化的方法，同时实现了显著更低的推理延迟。我们的代码已发布于https://github.com/git-disl/A4VL。