mlsanityData to 14 Oct 2026

Zero-shot 2D Grounding with Novel Affordance Types

arXiv 2608.08929 ↗First posted 9 Aug 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

2D affordance grounding aims to locate the region of an object that a human can interact with. Existing research focuses on recognizing affordance types seen during training and does not study models' ability to generalize to novel affordances, which is crucial for real-world applications. We propose the task of zero-shot 2D grounding with novel affordance types (NAT) and introduce the NAT benchmarks. We then propose AffordAnything, a training-free method that leverages segmentation cues, motivated by the strong correlation between affordance regions and object subparts. To further improve performance, we develop AffordAnything+, a trainable variant that learns to combine these cues. On the proposed AGD20K-NAT benchmark, our best model AffordAnything+ achieves a substantial improvement of 12.3% (absolute) in [email protected] over the SOTA affordance grounding method, OOAL.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 72 of its 89 references that are in the map. First posted on arXiv 9 Aug 2026.

Builds on

72 of 89 references.

References · 72

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2020Language Models are Few-Shot Learnerscited by 32,419
  4. 2023GPT-4 Technical Reportcited by 18,396
  5. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  6. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  7. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,332
  8. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,348
  9. 2024Grounded SAM: Assembling Open-World Models for Diverse Visual Taskscited by 929
  10. 2023VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Modelscited by 750
  11. 2021Extract Free Dense Labels from CLIPcited by 583
  12. 2023Florence-2: Advancing a Unified Representation for a Variety of Vision Taskscited by 410
38 more
  1. 2023Side Adapter Network for Open-Vocabulary Semantic Segmentationcited by 300
  2. 2023Affordances from Human Videos as a Versatile Representation for Roboticscited by 226
  3. 2021Decoupling Zero-Shot Semantic Segmentationcited by 199
  4. 2023Universal Instance Perception as Object Discovery and Retrievalcited by 192
  5. 2022ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentationcited by 178
  6. 2023SCLIP: Rethinking Self-Attention for Dense Vision-Language Inferencecited by 170
  7. 2023Towards Open Vocabulary Learning: A Surveycited by 166
  8. 2018Grounded Human-Object Interaction Hotspots from Videocited by 129
  9. 2017AffordanceNet: An End-to-End Deep Learning Approach for Object Affordance Detectioncited by 128
  10. 2024ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inferencecited by 113
  11. 2024ProxyCLIP: Proxy Attention Improves CLIP for Open-Vocabulary Segmentationcited by 111
  12. 2022Grounding Language with Visual Affordances over Unstructured Datacited by 101
  13. 2023SED: A Simple Encoder-Decoder for Open-Vocabulary Semantic Segmentationcited by 97
  14. 2023Language Embedded Radiance Fields for Zero-Shot Task-Oriented Graspingcited by 91
  15. 2023Grounding Everything: Emerging Localization Properties in Vision-Language Transformerscited by 91
  16. 2022Learning Affordance Grounding from Exocentric Imagescited by 82
  17. 2018Predicting Gaze in Egocentric Video by Learning Task-dependent Attention Transitioncited by 78
  18. 2017Learning to Act Properly: Predicting and Explaining Affordances from Imagescited by 74
  19. 2024Robo-ABC: Affordance Generalization Beyond Categories via Semantic Correspondence for Robot Manipulationcited by 71
  20. 2017SalGAN: Visual Saliency Prediction with Generative Adversarial Networkscited by 65
  21. 2020Learning Affordance Landscapes for Interaction Exploration in 3D Environmentscited by 65
  22. 2023LOCATE: Localize and Transfer Object Parts for Weakly Supervised Affordance Groundingcited by 64
  23. 2023Grounding 3D Object Affordance from 2D Interactions in Imagescited by 62
  24. 2022End-to-End Affordance Learning for Robotic Manipulationcited by 57
  25. 2016DeepGaze II: Reading fixations from deep features trained on object recognitioncited by 55
  26. 2024AffordanceLLM: Grounding Affordance from Vision Language Modelscited by 55
  27. 2016A Deep Multi-Level Network for Saliency Predictioncited by 54
  28. 2023One-Shot Open Affordance Learning with Foundation Modelscited by 45
  29. 2023Open-Vocabulary Affordance Detection in 3D Point Cloudscited by 41
  30. 2021One-Shot Object Affordance Detection in the Wildcited by 40
  31. 2023Affordance Grounding from Demonstration Video to Target Imagecited by 34
  32. 2024Learning Precise Affordances from Egocentric Videos for Robotic Manipulationcited by 34
  33. 2022One-Shot Transfer of Affordance Regions? AffCorrs!cited by 31
  34. 20253D-AffordanceLLM: Harnessing Large Language Models for Open-Vocabulary Affordance Detection in 3D Worldscited by 30
  35. 2024GREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance Groundingcited by 28
  36. 2025AffordDexGrasp: Open-set Language-guided Dexterous Grasp with Generalizable-Instructive Affordancecited by 27
  37. 2021Learning Visual Affordance Grounding from Demonstration Videoscited by 25
  38. 2024Learning 2D Invariant Affordance Knowledge for 3D Affordance Groundingcited by 22

50 most cited of 72 All 72 (JSON)

Built on by

Not cited yet.