DR$^2$Seg: Decomposed Two-Stage Rollouts for Efficient Reasoning Segmentation in Multimodal Large Language Models

Reasoning segmentation is an emerging vision-language task that requires reasoning over intricate text queries to precisely segment objects. However, existing methods typically suffer from overthinking, generating verbose reasoning chains that interfere with object localization in multimodal large language models (MLLMs). To address this issue, we propose DR$^2$Seg, a self-rewarding framework that improves both reasoning efficiency and segmentation accuracy without requiring extra thinking supervision. DR$^2$Seg employs a two-stage rollout strategy that decomposes reasoning segmentation into multimodal reasoning and referring segmentation. In the first stage, the model generates a self-contained description that explicitly specifies the target object. In the second stage, this description replaces the original complex query to verify its self-containment. Based on this design, two self-rewards are introduced to mitigate overthinking and the associated attention dispersion. Extensive experiments conducted on 3B and 7B variants of Qwen2.5-VL, as well as on both SAM2 and SAM3, demonstrate that DR$^2$Seg consistently improves reasoning efficiency and overall segmentation accuracy. The source code can be found at https://github.com/harrylin-hyl/DR2Seg.

We propose DR$^2$Seg, a simple yet effective self-reward framework that enhances both efficiency and segmentation accuracy using only the model’s intrinsic capability, without requiring extra MLLMs or supervision. DR$^2$Seg designs a two-stage rollout strategy that decouples multimodal reasoning and perception in MLLM for accurate segmentation, combined with a length-based self-reward to reduce redundant reasoning. Extensive experiments validate the effectiveness and generalization of DR$^2$Seg across MLLMs of varying scales and segmentation models, offering valuable insights into efficient reasoning perception.