mlsanityData to 7 Oct 2026

OV-Stitcher: A Global Context-Aware Framework for Training-Free Open-Vocabulary Semantic Segmentation

arXiv 2604.08110 ↗First posted 9 Apr 2026cs.CV cs.AI cs.LGcited by 0

On the map: Multimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP

Abstract (arXiv)

Training-free open-vocabulary semantic segmentation(TF-OVSS) has recently attracted attention for its ability to perform dense prediction by leveraging the pretrained knowledge of large vision and vision-language models, without requiring additional training. However, due to the limited input resolution of these pretrained encoders, existing TF-OVSS methods commonly adopt a sliding-window strategy that processes cropped sub-images independently. While effective for managing high-resolution inputs, this approach prevents global attention over the full image, leading to fragmented feature representations and limited contextual reasoning. We propose OV-Stitcher, a training-free framework that addresses this limitation by stitching fragmented sub-image features directly within the final encoder block. By reconstructing attention representations from fragmented sub-image features, OV-Stitcher enables global attention within the final encoder block, producing coherent context aggregation and spatially consistent, semantically aligned segmentation maps. Extensive evaluations across eight benchmarks demonstrate that OV-Stitcher establishes a scalable and effective solution for open-vocabulary segmentation, achieving a notable improvement in mean Intersection over Union(mIoU) from 48.7 to 50.7 compared with prior training-free baselines.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 56 of its 65 references that are in the map. First posted on arXiv 9 Apr 2026.

Builds on

56 of 65 references.

Closest ancestors · 9

Same direction, newest first

  1. 2025Mamba as a Bridge: Where Vision Foundation Models Meet Vision Language Models for Domain-Generalized Semantic Segmentation4 Apr
  2. 2024Distilling Spectral Graph for Object-Context Aware Open-Vocabulary Semantic Segmentation26 Nov
  3. 2024Self-Calibrated CLIP for Training-Free Open-Vocabulary Segmentation24 Nov
  4. 2024ResCLIP: Residual Attention for Training-free Dense Vision-language Inference24 Nov
  5. 2024ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inference17 Jul
  6. 2024Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation11 Jul
  7. 2023Rewrite Caption Semantics: Bridging Semantic Gaps for Language-Supervised Semantic Segmentation24 Sep
  8. 2023A Closer Look at the Explainability of Contrastive Language-Image Pre-training12 Apr
  9. 2023Side Adapter Network for Open-Vocabulary Semantic Segmentation23 Feb

Other references · 47

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2024The Llama 3 Herd of Modelscited by 14,192
  4. 2023Segment Anythingcited by 8,158
  5. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  6. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  7. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  8. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  9. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
  10. 2024SAM 2: Segment Anything in Images and Videoscited by 2,784
  11. 2023FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioningcited by 2,551
  12. 2025DINOv3cited by 1,297
35 more
  1. 2022Reproducible scaling laws for contrastive language-image learningcited by 1,163
  2. 2021SimMIM: A Simple Framework for Masked Image Modelingcited by 1,147
  3. 2023Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecturecited by 982
  4. 2016COCO-Stuff: Thing and Stuff Classes in Contextcited by 858
  5. 2023Vision Transformers Need Registerscited by 844
  6. 2021ImageNet-21K Pretraining for the Massescited by 607
  7. 2021Extract Free Dense Labels from CLIPcited by 580
  8. 2022Open-Vocabulary Semantic Segmentation with Mask-adapted CLIPcited by 560
  9. 2023Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistlescited by 287
  10. 2023Data Filtering Networkscited by 247
  11. 2023CAT-Seg: Cost Aggregation for Open-Vocabulary Semantic Segmentationcited by 215
  12. 2023Convolutions Die Hard: Open-Vocabulary Segmentation with Single Frozen Convolutional CLIPcited by 199
  13. 2023SCLIP: Rethinking Self-Attention for Dense Vision-Language Inferencecited by 169
  14. 2022SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary Semantic Segmentationcited by 162
  15. 2022Learning to Generate Text-grounded Mask for Open-world Semantic Segmentation from Only Image-Text Pairscited by 130
  16. 2024ProxyCLIP: Proxy Attention Improves CLIP for Open-Vocabulary Segmentationcited by 110
  17. 2023SED: A Simple Encoder-Decoder for Open-Vocabulary Semantic Segmentationcited by 95
  18. 2023CLIP-DINOiser: Teaching CLIP a few DINO tricks for open-vocabulary semantic segmentationcited by 94
  19. 2023Grounding Everything: Emerging Localization Properties in Vision-Language Transformerscited by 89
  20. 2023Stronger, Fewer, & Superior: Harnessing Vision Foundation Models for Domain Generalized Semantic Segmentationcited by 82
  21. 2024Pay Attention to Your Neighbours: Training-Free Open-Vocabulary Semantic Segmentationcited by 76
  22. 2023CLIP as RNN: Segment Countless Visual Concepts without Training Endeavorcited by 69
  23. 2023Open-Vocabulary Segmentation with Semantic-Assisted Calibrationcited by 49
  24. 2024In Defense of Lazy Visual Grounding for Open-Vocabulary Semantic Segmentationcited by 38
  25. 2024Harnessing Vision Foundation Models for High-Performance, Training-Free Open Vocabulary Segmentationcited by 38
  26. 2024CorrCLIP: Reconstructing Patch Correlations in CLIP for Open-Vocabulary Semantic Segmentationcited by 33
  27. 2025SAM-R1: Leveraging SAM for Reward Feedback in Multimodal Segmentation via Reinforcement Learningcited by 31
  28. 2023Uncovering Prototypical Knowledge for Weakly Open-Vocabulary Semantic Segmentationcited by 27
  29. 2023Distilling Self-Supervised Vision Transformers for Weakly-Supervised Few-Shot Classification & Segmentationcited by 24
  30. 2024SoMA: Singular Value Decomposed Minor Components Adaptation for Domain Generalizable Representation Learningcited by 22
  31. 2024Image-Text Co-Decomposition for Text-Supervised Semantic Segmentationcited by 21
  32. 2023Window Attention is Bugged: How not to Interpolate Position Embeddingscited by 18
  33. 2024Towards Generalizable Scene Change Detectioncited by 17
  34. 2024USE: Universal Segment Embeddings for Open-Vocabulary Image Segmentationcited by 14
  35. 2025TextRegion: Text-Aligned Region Tokens from Frozen Image-Text Modelscited by 7

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Emergent Open-Vocabulary Semantic Segmentation from Off-the-shelf Vision-Language Models
  2. 2025Efficient Redundancy Reduction for Open-Vocabulary Semantic Segmentation
  3. 2024Transferable and Principled Efficiency for Open-Vocabulary Segmentation
  4. 2026dinov3.seg: Open-Vocabulary Semantic Segmentation with DINOv3
  5. 2025Training-Free Class Purification for Open-Vocabulary Semantic Segmentation
  6. 2025OV-COAST: Cost Aggregation with Optimal Transport for Open-Vocabulary Semantic Segmentation
  7. 2023Global Knowledge Calibration for Fast Open-Vocabulary Segmentation
  8. 2026Retrieve and Segment: Are a Few Examples Enough to Bridge the Supervision Gap in Open-Vocabulary Segmentation?