arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.04002cs.CVcs.AI
收藏

通过锚定描述实现的判别感知用于推理分割

Discriminative Perception via Anchored Description for Reasoning Segmentation

Northwestern Polytechnical University(西北工业大学)

Tao Yang, Qing Zhou, Yanliang Li, Qi Wang

AI总结:

本文提出DPAD方法,通过生成描述性标题来增强推理分割的判别能力,提升推理链的聚焦性和效率。

摘要:

推理分割越来越多地采用强化学习来生成解释性推理链,以指导多模态大语言模型。虽然这些几何奖励主要局限于指导最终定位,但它们无法判断推理过程是否仍然锚定在所指的区域或偏离到无关的上下文。缺乏这种判别指导,模型的推理往往会退化为无焦点且冗长的链条,最终无法在复杂场景中澄清和感知目标。这表明需要补充RL目标以实现判别感知,即能够主动区分目标与其上下文的能力。为此,我们提出DPAD,迫使模型生成所指对象的描述性标题,然后通过将标题的语义相关性与更广泛的上下文进行对比来显式判别。通过优化这种判别能力,模型被迫专注于目标的独特属性,从而产生更集中和高效的推理链。描述性标题还充当了可解释性理由,与分割对齐。在基准测试上的实验确认了我们方法的有效性,带来了显著的性能提升,ReasonSeg的cIoU增加了3.09%,推理链长度减少了约42%。代码可在https://github.com/mrazhou/DPAD上获得。

英文摘要:

Reasoning segmentation increasingly employs reinforcement learning to generate explanatory reasoning chains that guide Multimodal Large Language Models. While these geometric rewards are primarily confined to guiding the final localization, they are incapable of discriminating whether the reasoning process remains anchored on the referred region or strays into irrelevant context. Lacking this discriminative guidance, the model's reasoning often devolves into unfocused and verbose chains that ultimately fail to disambiguate and perceive the target in complex scenes. This suggests a need to complement the RL objective with Discriminative Perception, an ability to actively distinguish a target from its context. To realize this, we propose DPAD to compel the model to generate a descriptive caption of the referred object, which is then used to explicitly discriminate by contrasting the caption's semantic relevance to the referred object against the wider context. By optimizing for this discriminative capability, the model is forced to focus on the unique attributes of the target, leading to a more converged and efficient reasoning chain. The descriptive caption also serves as an interpretability rationale that aligns with the segmentation. Experiments on the benchmarks confirm the validity of our approach, delivering substantial performance gains, with the cIoU on ReasonSeg increasing by 3.09% and the reasoning chain length decreasing by approximately 42%. Code is available at https://github.com/mrazhou/DPAD

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑