Qwen3-VL-Seg: Unlocking Open-World Referring Segmentation with Vision-Language Grounding

• URL: http://arxiv.org/abs/2605.07141v1
• Date: Fri, 08 May 2026 02:20:40 GMT
• Title: Qwen3-VL-Seg: Unlocking Open-World Referring Segmentation with Vision-Language Grounding
• Abstract summary: We present Qwen3-VL-Seg, a parameter-efficient framework that treats the MLLM-predicted box as a semantically grounded structural prior.<n>At its core, a lightweight box-guided mask decoder combines multi-scale spatial feature injection, spatial-semantic query construction, box-guided high-resolution pixel fusion.<n>Experiments on referring expression segmentation, visual grounding, and ORS-Bench show that Qwen3-VL-Seg performs strongly across closed-set and open-world settings.
• Score: 26.30521907946121
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Open-world referring segmentation requires grounding unconstrained language expressions to precise pixel-level regions. Existing multimodal large language models (MLLMs) exhibit strong open-world visual grounding, but their outputs remain limited to sparse bounding-box coordinates and are insufficient for dense visual prediction. Recent MLLM-based segmentation methods either directly predict sparse contour coordinates, struggling to reconstruct continuous object boundaries, or rely on external segmentation foundation models such as the Segment Anything Model (SAM), introducing substantial architectural and deployment overhead. We present Qwen3-VL-Seg, a parameter-efficient framework that treats the MLLM-predicted box as a semantically grounded structural prior and decodes it into pixel-level referring segmentation. At its core, a lightweight box-guided mask decoder combines multi-scale spatial feature injection, spatial-semantic query construction, box-guided high-resolution pixel fusion, and iterative mask-aware query refinement, introducing only 17M parameters (about 0.4\% of the base model). For scalable open-world training, we construct SA1B-ORS, an SA-1B-derived dataset with two subsets: SA1B-CoRS (category-oriented samples) and SA1B-DeRS (descriptive, instance-specific samples). For evaluation, we curate ORS-Bench, a manually screened benchmark with in-distribution and out-of-distribution subsets covering diverse referring expression types. Extensive experiments on referring expression segmentation, visual grounding, and ORS-Bench show that Qwen3-VL-Seg performs strongly across closed-set and open-world settings, with clear advantages on language-intensive instructions and strong out-of-distribution generalization. Evaluations on general multimodal benchmarks further show that the model broadly preserves general-purpose multimodal competence after segmentation-oriented adaptation.

Related papers

• Seg-Agent: Test-Time Multimodal Reasoning for Training-Free Language-Guided Segmentation [52.8308168727975]
  Seg-Agent is a completely training-free framework that pioneers Explicit Multimodal Chain-of-Reasoning.<n>Our approach constructs an interactive visual reasoning loop comprising three stages: generation, selection, and refinement.<n>Various-LangSeg is a novel benchmark covering explicit semantic, generic object, and reasoning-guided segmentation tasks.
  arXiv  Detail & Related papers  (2026-05-13T03:36:44Z)
• PCA-Seg: Revisiting Cost Aggregation for Open-Vocabulary Semantic and Part Segmentation [58.1914505657064]
  This paper proposes a simple yet effective parallel cost aggregation (PCA-Seg) paradigm to alleviate the challenge of knowledge interference between class-level semantics and spatial context.<n>Experiments on eight benchmarks show that each parallel block in PCA-Seg adds merely 0.35M parameters while achieving state-of-the-art OSPS performance.
  arXiv  Detail & Related papers  (2026-03-18T09:26:43Z)
• Segment and Matte Anything in a Unified Model [5.8874968768571625]
  Segment Anything (SAM) has recently pushed the boundaries of segmentation by demonstrating zero-shot generalization and flexible prompting.<n>We introduce Segment And Matte Anything (SAMA), a lightweight extension of SAM that delivers high-quality interactive image segmentation and matting.
  arXiv  Detail & Related papers  (2026-01-17T19:43:10Z)
• Bridging Semantics and Geometry: A Decoupled LVLM-SAM Framework for Reasoning Segmentation in Remote Sensing [8.731693840957716]
  Think2Seg-RS is a framework that trains an LVLM prompter to control a frozen Segment Anything Model (SAM) via structured geometric prompts.<n>The framework achieves state-of-the-art performance on the EarthReason dataset.<n> compact segmenters outperform larger ones under semantic-level supervision, and that negative prompts are ineffective in heterogeneous aerial backgrounds.
  arXiv  Detail & Related papers  (2025-12-22T11:46:42Z)
• FineRS: Fine-grained Reasoning and Segmentation of Small Objects with Reinforcement Learning [62.11389260206383]
  textscFineRS is a two-stage MLLM-based reinforcement learning framework for segmenting extremely small objects.<n>We present textscFineRS-4k, a new dataset for evaluating MLLMs on attribute-level reasoning and pixel-level segmentation on subtle, small-scale targets.
  arXiv  Detail & Related papers  (2025-10-24T10:14:17Z)
• X-SAM: From Segment Anything to Any Segmentation [63.79182974315084]
  Large Language Models (LLMs) demonstrate strong capabilities in broad knowledge representation, yet they are inherently deficient in pixel-level perceptual understanding.<n>We present X-SAM, a streamlined Multimodal Large Language Model framework that extends the segmentation paradigm from textitsegment anything to textitany segmentation.<n>We propose a new segmentation task, termed Visual GrounDed (VGD) segmentation, which segments all instance objects with interactive visual prompts and empowers MLLMs with visual grounded, pixel-wise interpretative capabilities.
  arXiv  Detail & Related papers  (2025-08-06T17:19:10Z)
• LIRA: Inferring Segmentation in Large Multi-modal Models with Local Interleaved Region Assistance [54.683384204063934]
  Large multi-modal models (LMMs) struggle with inaccurate segmentation and hallucinated comprehension.<n>We propose LIRA, a framework that capitalizes on the complementary relationship between visual comprehension and segmentation.<n>LIRA achieves state-of-the-art performance in both segmentation and comprehension tasks.
  arXiv  Detail & Related papers  (2025-07-08T07:46:26Z)
• OpenWorldSAM: Extending SAM2 for Universal Image Segmentation with Language Prompts [20.327695503392274]
  We present OpenWorldSAM, a framework that extends the prompt-driven Segment Anything Model v2 (SAM2) to open-vocabulary scenarios.<n>OpenWorldSAM supports a diverse range of prompts, including category-level and sentence-level language descriptions.<n>We enhance the model's spatial understanding through novel positional tie-breaker embeddings and cross-attention layers.
  arXiv  Detail & Related papers  (2025-07-07T19:16:22Z)
• SAM-MIL: A Spatial Contextual Aware Multiple Instance Learning Approach for Whole Slide Image Classification [9.69491390062406]
  We propose a novel MIL framework, named SAM-MIL, that emphasizes spatial contextual awareness and explicitly incorporates spatial context. Our approach includes the design of group feature extraction based on spatial context and a SAM-Guided Group Masking strategy. Experimental results on the CAMELYON-16 and TCGA Lung Cancer datasets demonstrate that our proposed SAM-MIL model outperforms existing mainstream methods in WSIs classification.
  arXiv  Detail & Related papers  (2024-07-25T01:12:48Z)
• PosSAM: Panoptic Open-vocabulary Segment Anything [58.72494640363136]
  PosSAM is an open-vocabulary panoptic segmentation model that unifies the strengths of the Segment Anything Model (SAM) with the vision-native CLIP model in an end-to-end framework. We introduce a Mask-Aware Selective Ensembling (MASE) algorithm that adaptively enhances the quality of generated masks and boosts the performance of open-vocabulary classification during inference for each image.
  arXiv  Detail & Related papers  (2024-03-14T17:55:03Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.