Reasoning via Implicit Self-supervised Emergence for Instruction Segmentation
Qing Zhou, Lichang Yang, Yuyu Jia, Junyu Gao, Weiping Ni, Junzheng Wu, Qi Wang
Abstract
We challenge the assumption that complex instruction-guided segmentation tasks necessitate equally complex and explicit supervision. This paper introduces RISE (Reasoning via Implicit Self-supervised Emergence), a framework that learns intricate compositional reasoning, spanning spatial relations to world knowledge, without a single ground-truth mask. To achieve this, RISE employs reinforcement learning with GRPO guided by a single, strikingly simple reward: the semantic alignment score between the textual instruction and the predicted image region. Our primary discovery is the implicit emergence of a high-quality chain-of-thought process from this minimalist signal. Within a structured format, the model autonomously learns to understand instructions by accessing its latent knowledge, inferring spatial relationships—capabilities inherent in its architecture but unlocked by our simple objective. Remarkably, our emergent reasoning yields highly competitive results: RISE achieves 58.7 gIoU on the ReasonSeg benchmark, on par with methods using geometric rewards. Furthermore, we show extreme data efficiency: a variant trained on only 2,000 ImageNet-label pairs establishes a new state-of-the-art for annotation-free referring segmentation with 79.6 cIoU on RefCOCO.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- CoT-Valve: Length-Compressible Chain-of-Thought Tuning cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- One Token to Seg Them All: Language Instructed Reasoning Segmentation in Videos cites LAVT: Language-Aware Vision Transformer for Referring Image Segmentation
- LAVT: Language-Aware Vision Transformer for Referring Image Segmentation cites Learning Transferable Visual Models From Natural Language Supervision
Builds on13
- Learning Transferable Visual Models From Natural Language SupervisionAlec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh et al.ICML 2021 · 47,906 citations
- Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma et al.NeurIPS 2022 · 22,562 citations
- LAVT: Language-Aware Vision Transformer for Referring Image SegmentationZhao Yang, Jiaqi Wang, Yansong Tang, Kai Chen et al.CVPR 2022 · 319 citations
- CoT-Valve: Length-Compressible Chain-of-Thought TuningXinyin Ma, Guangnian Wan, Runpeng Yu, Gongfan Fang et al.ACL 2025 · 162 citations
- One Token to Seg Them All: Language Instructed Reasoning Segmentation in VideosZechen Bai, Tong He, Haiyang Mei, Pichao Wang et al.NeurIPS 2024 · 147 citations
Related papers
- VideoSeg-R1: Reasoning Video Object Segmentation via Reinforcement LearningZishan Xu, Yifu Guo, Yuquan Lu, Fengyu Yang et al.AAAI 2026
- RISE: Enhancing VLM Image Annotation with Self-Supervised ReasoningSuhang Hu, Wei Hu, Yuhang Su, Fan ZhangCVPR Findings 2026
- SegCompass: Exploring Interpretable Alignment with Sparse Autoencoders for Enhanced Reasoning SegmentationZhenyu Lu, Liupeng Li, Jinpeng Wang, Haoqian Kang et al.CVPR 2026 · 3 citations
- Hybrid Global-Local Representation with Augmented Spatial Guidance for Zero-Shot Referring Image SegmentationTing Liu, Siyuan LiCVPR 2025
- LENS: Learning to Segment Anything with Unified Reinforced ReasoningLianghui Zhu, Bin Ouyang, Yuxuan Zhang, Tianheng Cheng et al.AAAI 2026 · 7 citations
