mlsanityData to 14 Oct 2026

Open-Vocabulary Semantic Segmentation in Remote Sensing via Hierarchical Attention Masking and Model Composition

arXiv 2602.23869 ↗First posted 27 Feb 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

In this paper, we propose ReSeg-CLIP, a new training-free Open-Vocabulary Semantic Segmentation method for remote sensing data. To compensate for the problems of vision language models, such as CLIP in semantic segmentation caused by inappropriate interactions within the self-attention layers, we introduce a hierarchical scheme utilizing masks generated by SAM to constrain the interactions at multiple scales. We also present a model composition approach that averages the parameters of multiple RS-specific CLIP variants, taking advantage of a new weighting scheme that evaluates representational quality using varying text prompts. Our method achieves state-of-the-art results across three RS benchmarks without additional training.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 39 of its 49 references that are in the map. First posted on arXiv 27 Feb 2026.

Builds on

39 of 49 references.

References · 39

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  4. 2023Segment Anythingcited by 8,234
  5. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,864
  6. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,209
  7. 2022Editing Models with Task Arithmeticcited by 1,120
  8. 2021Robust fine-tuning of zero-shot modelscited by 797
  9. 2022Language-driven Semantic Segmentationcited by 717
  10. 2022GroupViT: Semantic Segmentation Emerges from Text Supervisioncited by 531
  11. 2021Scaling Open-Vocabulary Image Segmentation with Image-Level Labelscited by 473
  12. 2023LoraHub: Efficient Cross-Task Generalization via Dynamic LoRA Compositioncited by 364
27 more
  1. 2023RemoteCLIP: A Vision Language Foundation Model for Remote Sensingcited by 363
  2. 2019Lookahead Optimizer: k steps forward, 1 step backcited by 336
  3. 2019BigEarthNet: A Large-Scale Benchmark Archive For Remote Sensing Image Understandingcited by 234
  4. 2021A Simple Baseline for Open-Vocabulary Semantic Segmentation with Pre-trained Vision-language Modelcited by 223
  5. 2023Convolutions Die Hard: Open-Vocabulary Segmentation with Single Frozen Convolutional CLIPcited by 201
  6. 2021Decoupling Zero-Shot Semantic Segmentationcited by 199
  7. 2022MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image Pretrainingcited by 187
  8. 2024Mixture of LoRA Expertscited by 181
  9. 2023SCLIP: Rethinking Self-Attention for Dense Vision-Language Inferencecited by 170
  10. 2022SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary Semantic Segmentationcited by 164
  11. 2023Composing Parameter-Efficient Modules with Arithmetic Operationscited by 156
  12. 2023RS5M and GeoRSCLIP: A Large Scale Vision-Language Dataset and A Large Vision-Language Model for Remote Sensingcited by 151
  13. 2023SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensingcited by 125
  14. 2024Twin-Merging: Dynamic Integration of Modular Expertise in Model Mergingcited by 116
  15. 2024ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inferencecited by 113
  16. 2023A Closer Look at the Explainability of Contrastive Language-Image Pre-trainingcited by 96
  17. 2023Grounding Everything: Emerging Localization Properties in Vision-Language Transformerscited by 91
  18. 2023AdapterSoup: Weight Averaging to Improve Generalization of Pretrained Language Modelscited by 84
  19. 2023Model Merging by Uncertainty-Based Gradient Matchingcited by 77
  20. 2024Pay Attention to Your Neighbours: Training-Free Open-Vocabulary Semantic Segmentationcited by 77
  21. 2024Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentationcited by 61
  22. 2024Knowledge Composition using Task Vectors with Learned Anisotropic Scalingcited by 37
  23. 2024DAM: Dynamic Adapter Merging for Continual Video QA Learningcited by 19
  24. 2024Towards Open-Vocabulary Remote Sensing Image Semantic Segmentationcited by 19
  25. 2024Continual Learning with Weight Interpolationcited by 13
  26. 2024Open-Vocabulary Remote Sensing Image Semantic Segmentationcited by 11
  27. 2025AerOSeg: Harnessing SAM for Open-Vocabulary Segmentation in Remote Sensing Imagescited by 9

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026SatOV: Restoring Spatial Priors for Training-Free Open-Vocabulary Segmentation in Remote Sensing Imagery
  2. 2025Annotation-Free Open-Vocabulary Segmentation for Remote-Sensing Images
  3. 2026ConInfer: Context-Aware Inference for Training-Free Open-Vocabulary Remote Sensing Segmentation
  4. 2026Prompt-Calibrated SAM 3 for Open-Vocabulary Remote Sensing Semantic Segmentation
  5. 2026Enabling Training-Free Text-Based Remote Sensing Segmentation
  6. 2026Geospatial-Reasoning-Driven Vocabulary-Agnostic Remote Sensing Semantic Segmentation
  7. 2025UniGeoSeg: Towards Unified Open-World Segmentation for Geospatial Scenes
  8. 2025DGL-RSIS: Decoupling Global Spatial Context and Local Class Semantics for Training-Free Remote Sensing Image Segmentation