Shadowgraph
TrendingAbout
Sign in

hq-King/Awesome-Affordance-Learning

229

68 commits

updated Sep 10, 2026

See the code

README

From Passive Perception to Active Interaction: A Survey of
Affordance Learning for Embodied AI

Awesome License Last Commit GitHub Stars

Gen Li1,#,‡, Hanqing Wang2,#, Jingliang Li1,#, Yifan Han3,#, Jindou Jia1, Tao Lin3, Yuanzhe Liu4, Yutong Wang5, Bo Zhao3, Fangqiang Ding2, Anh Nguyen6, Laura Sevilla-Lara7, Huazhe Xu8, Gregory S. Chirikjian9,
Marc Pollefeys10, Oier Mees10, Hui Xiong2,†, Jianfei Yang1,†

1Nanyang Technological University    2HKUST(GZ)    3SJTU 4UIUC    5The University of Sydney   
6University of Liverpool    7The University of Edinburgh 8MBZUAI    9Tsinghua University    10ETH Zurich

#Equal Contribution    ‡Project Lead    †Corresponding Author

A curated collection of papers on affordance learning for embodied AI.

🧭 Exploring Embodied AI and Embodied perception? We hope this collection proves useful in your journey. If you'd like to support the project, feel free to ⭐️ the repo and share it with your peers. Contributions are warmly welcome!

📖 Contents

🔥 News

📢 This list is actively maintained, and community contributions are always appreciated!
Feel free to open a pull request if you find any relevant papers.

  • [2025-05] 🎉 This repository was launched to curate a comprehensive list of affordance-learning research.

🌟 Introduction

This repository accompanies the survey From Passive Perception to Active Interaction: A Survey of Affordance Learning for Embodied AI and maintains a curated collection of papers, datasets, and benchmarks.

As robots and embodied agents move into real-world applications, they must understand not only what objects are, but also where, why, and how they can interact with them. Following the survey, the list is organized around three complementary questions:

  • Affordance Perception: Where or which region affords an interaction?
  • Affordance Reasoning: Which affordance is relevant under the current task, environment, and constraints?
  • Affordance-Guided Action: How can an affordance be converted into an executable action?

A paper may be cross-referenced when the survey discusses it in more than one role (for example, both perception and reasoning). A separate section collects datasets and benchmarks. Venue labels use the formally published version whenever one is available; otherwise the first public preprint is marked arXiv.

🧭 Taxonomy

The taxonomy follows the survey's functional pipeline rather than only model architecture. It expands each primary category into second- and third-level categories. Cross-listing is intentional when a method contributes to multiple stages or perspectives.

Primary CategorySecondary CategoryThird-Level CategoriesMain Distinction
Affordance PerceptionVisual affordance perceptionObject-centric grounding; scene-level grounding; weakly supervised perceptionGrounds masks, heatmaps, or keypoints on the 2D image plane.
Affordance PerceptionSpatial affordance perception3D object grounding; 3D scene groundingGrounds functional regions or geometric structures directly in metric 3D space.
Affordance PerceptionInteraction-driven affordance perceptionDemonstration/HOI video; HOI image; interaction-conditioned 3D; interaction-grounded scene perceptionDerives supervision or context from observed human-object interactions.
Affordance PerceptionGeneralizable affordance perceptionExample-based transfer; open-set groundingTransfers to novel objects, labels, queries, or interaction contexts.
Affordance ReasoningRelation-based reasoningProbabilistic/semantic relations; object-pair and scene context; manipulation graphsInfers affordances through explicit relations among objects, actions, agents, and scene context.
Affordance ReasoningLanguage-centric reasoningGrounded skill selection; integrated LLM/MLLM prediction; modular semantic-to-spatial grounding; sequential reasoningUses language to interpret intent and select or ground task-relevant affordances.
Affordance ReasoningAgentic reasoningPredefined workflows; adaptive workflowsUses planning, memory, verification, or iterative tool/model calls over multiple steps.
Affordance-Guided ActionHierarchical affordance-to-actionPrimitive-based execution; retrieval-based transfer; planner-based executionPredicts affordances first and then converts them into actions through a separate execution module.
Affordance-Guided ActionAffordance-integrated policy learningAffordance as explicit input; implicit representation; optimization signalIntegrates affordance information directly into policy representation, input, or optimization.

📄 Paper List

The Venue/Date column prioritizes the formal venue and publication year. Papers without a confirmed venue are labeled arXiv. The Name column records the method or system name explicitly introduced by the authors, including names stated only in the abstract or main text rather than in the title; - means that no explicit method name has been verified and no acronym is inferred from the title.

👁️ Affordance Perception

Venue/DateNameTitlePaperScopePerception Subcategory
IJRR 2013-Learning Human Activities and Object Affordances from RGB-D VideosPaperSceneInteraction-driven
ICRA 2015UMDAffordance Detection of Tool Parts from Geometric FeaturesPaperObjectVisual · Spatial
ECCV 2016-A Multi-scale CNN for Affordance Segmentation in RGB ImagesPaperSceneVisual
IROS 2016-Detecting Object Affordances with Convolutional Neural NetworksPaperObjectVisual
CVPR 2017-Weakly Supervised Affordance DetectionPaperObjectVisual
IROS 2017-Object-Based Affordances Detection with Convolutional Neural Networks and Dense Conditional Random FieldsPaperObjectVisual
ICCVW 2017-Adaptive Binarization for Weakly Supervised Affordance SegmentationPaperObjectVisual
ICCVW 2017-Learning to Segment AffordancesPaperObject/SceneVisual
Humanoids 2017-Affordance Detection for Task-Specific Grasping Using Deep LearningPaperObjectVisual
ICRA 2018AffordanceNetAffordanceNet: An End-to-End Deep Learning Approach for Object Affordance DetectionPaperObjectVisual
CVPR 2018Demo2VecDemo2Vec: Reasoning Object Affordances from Online VideosPaperObjectVisual · Interaction-driven
ECCV 2018Gaze + ActionIn the Eye of Beholder: Joint Learning of Gaze and Actions in First Person VideoPaperSceneInteraction-driven
Autonomous Robots 2019-Towards Affordance Detection for Robot Manipulation Using Affordance for Parts and Parts for AffordancePaperObjectVisual
RAS 2019-Context-Based Affordance Segmentation from 2D Images for Robot ActionsPaperObjectVisual
RA-L with ICRA 2019AffordanceLearnLearning Affordance Segmentation for Real-world Robotic Manipulation via Synthetic ImagesPaperObjectVisual
arXiv 2019AffContextRecognizing Object Affordances to Support Scene Reasoning for Manipulation TasksPaperObjectVisual
ICCV 2019Interaction HotspotsGrounded Human-Object Interaction Hotspots From VideoPaperObjectVisual · Interaction-driven
CVPR 2020Ego-TopoEgo-Topo: Environment Affordances from Egocentric VideoPaperSceneInteraction-driven
NeurIPS 2020IntExpLearning Affordance Landscapes for Interaction Exploration in 3D EnvironmentsPaperSceneSpatial
Neural Comput. Appl. 2020.09-Object Affordance Detection with Relationship-Aware NetworkPaperObjectVisual
RA-L with ICRA2021AffKpAn Affordance Keypoint Detection Network for Robot ManipulationPaperObjectVisual
ACM ICDAR 2021ST-HOIST-HOI: A Spatial-Temporal Baseline for Human-Object Interaction Detection in VideosPaperSceneInteraction-driven
CVPR 20213D AffordanceNet3D AffordanceNet: A Benchmark for Visual Object Affordance UnderstandingPaperObjectSpatial
CoRL 2021O2O-AffordO2O-Afford: Annotation-Free Large-Scale Object-Object Affordance LearningPaperObject-pairSpatial
Neurocomputing 2021-Visual Affordance Detection Using an Efficient Attention Convolutional Neural NetworkPaperObjectVisual
IJCAI 2021OS-ADOne-Shot Affordance DetectionPaperObjectVisual · Generalizable
Humanoids 2022ACFManipulation-Oriented Object Perception in Clutter through Affordance Coordinate FramesPaperObjectSpatial
ECCVW 2022PartAffordPartAfford: Part-level Affordance Discovery from 3D ObjectsPaperObjectSpatial
CVPR 2022OCTJoint Hand Motion and Interaction Hotspots Prediction from Egocentric VideosPaperObjectVisual · Interaction-driven
CVPR 2022Cross-view-AGLearning Affordance Grounding from Exocentric ImagesPaperObjectVisual · Interaction-driven
CoRL 2022AffCorrsOne-Shot Transfer of Affordance Regions? AffCorrs!PaperObjectVisual · Generalizable
Neural Comput. Appl. 2022BPNObject Affordance Detection with Boundary-Preserving Network for Robotic Manipulation TasksPaperObjectVisual
IJCV 2022OSAD-NetOne-Shot Object Affordance Detection in the WildPaperObjectVisual · Generalizable
WACV 2023-Fine-Grained Affordance Annotation for Egocentric Hand-Object Interaction VideosPaperObjectInteraction-driven
arXiv 2023STRAPSTRAP: Structured Object Affordance Segmentation with Point SupervisionPaperObjectVisual
CVPR 2023AfformerAffordance Grounding From Demonstration Video To Target ImagePaperObjectInteraction-driven
CVPR 2023Leverage Interactive Affinity for Affordance LearningPaperObjectVisual · Interaction-driven
CVPR 2023LOCATELOCATE: Localize and Transfer Object Parts for Weakly Supervised Affordance GroundingPaperObjectVisual · Interaction-driven
NeurIPS 2023Where2ExploreWhere2Explore: Few-shot Affordance Learning for Unseen Novel Categories of Articulated ObjectsPaperObjectSpatial · Generalizable
ICIP 2023-A Large Scale Multi-View RGBD Visual Affordance Learning DatasetPaperObjectVisual
IROS 2023VATHierarchical Transformer for Visual Affordance Understanding using a Large-scale DatasetPaperObjectVisual
ICCV 2023AffordPoseAffordPose: A Large-Scale Dataset of Hand-Object Interactions with Affordance-Driven Hand PosePaperObjectSpatial · Interaction-driven
IROS 2023HANDALHANDAL: A Dataset of Real-World Manipulable Object Categories with Pose Annotations, Affordances, and ReconstructionsPaperObjectVisual · Spatial
ICCV 2023IAGNetGrounding 3D Object Affordance from 2D Interactions in ImagesPaperObjectSpatial · Interaction-driven
ICCV 2023-Understanding 3D Object Interaction from a Single ImagePaperSceneSpatial
ICCV 2023EPIC-AffMulti-label Affordance Mapping from Egocentric VisionPaperSceneSpatial · Interaction-driven
IROS 2023OpenADOpen-vocabulary Affordance Detection in 3D Point CloudsPaperObjectSpatial · Generalizable
ICRA 2024-Open-Vocabulary Affordance Detection using Knowledge Distillation and Text-Point CorrelationPaperObjectSpatial · Generalizable
CVPR 2024LEMONLEMON: Learning 3D Human-Object Interaction Relation from 2D ImagesPaperObjectSpatial · Interaction-driven
AAAI 2024WSMAWeakly Supervised Multimodal Affordance Grounding for Egocentric ImagesPaperObjectVisual · Interaction-driven
arXiv 2024-Text-driven Affordance Learning from Egocentric VisionPaperObjectInteraction-driven
ICRA 20243DAPNetLanguage-Conditioned Affordance-Pose Detection in 3D Point CloudsPaperObjectSpatial · Generalizable
CVPR 2024OOALOne-Shot Open Affordance Learning with Foundation ModelsPaperObjectVisual · Generalizable
CVPR 2024LASOLASO: Language-guided Affordance Segmentation on 3D ObjectPaperObjectSpatial · Generalizable
CVPRW 2024LGAfford-NetLGAfford-Net: A Local Geometry Aware Affordance Detection Network for 3D Point CloudsPaperObjectSpatial
CVPR 2024SceneFun3DSceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D ScenesPaperSceneSpatial
arXiv 2024Ego-SAGGrounding 3D Scene Affordance From Egocentric InteractionsPaperSceneInteraction-driven · Spatial
ECCV 2024INTRAINTRA: Interaction Relationship-aware Weakly Supervised Affordance GroundingPaperObjectVisual · Interaction-driven
ECCV 2024ComABeyond the Contact: Discovering Comprehensive Affordance for 3D Objects from Pre-trained 2D Diffusion ModelsPaperObjectSpatial
IROS 2024FAKP-Net3D Affordance Keypoint Detection for Robotic ManipulationPaperObjectSpatial
ICCV 2025GLANCEGLANCE: Intermediate Connectors and Geometric Priors for Language-Guided Affordance Segmentation on Unseen Object CategoriesPaperObjectSpatial · Generalizable
AAAI 2025MIFAGLearning 2D Invariant Affordance Knowledge for 3D Affordance GroundingPaperObjectSpatial · Interaction-driven
CVPR 2025GEALGEAL: Generalizable 3D Affordance Learning with Cross-Modal ConsistencyPaperObjectSpatial · Generalizable
AAAI 2025MaskPromptMaskPrompt: Open-Vocabulary Affordance Segmentation with Object Shape Mask PromptsPaperObjectVisual · Generalizable
arXiv 2025AffordanceSAMAffordanceSAM: Segment Anything Once More in Affordance GroundingPaperObjectVisual · Generalizable
CVPR 2025IAAOIAAO: Interactive Affordance Learning for Articulated Objects in 3D EnvironmentsPaperObjectSpatial
arXiv 2025-Interpretable Affordance Detection on 3D Point Clouds with Probabilistic PrototypesPaperObjectSpatial
arXiv 2025-Weakly-Supervised Affordance Grounding Guided by Part-Level Semantic PriorsPaperObjectVisual
ICRA 2025UADUAD: Unsupervised Affordance Distillation for Generalization in Robotic ManipulationPaperObjectVisual · Generalizable
CVPR 2025Reasoning MambaReasoning Mamba: Hypergraph-Guided Region Relation Calculating for Weakly Supervised Affordance GroundingPaperObjectVisual
IROS 2025OS-AGDOOne-Shot Affordance Grounding of Deformable Objects in Egocentric Organizing ScenesPaperObjectVisual · Generalizable
IROS 2025BiT-AlignResource-Efficient Affordance Grounding with Complementary Depth and Semantic PromptsPaperObjectVisual
ICCV 2025LoopTransClosed-Loop Transfer for Weakly-supervised Affordance GroundingPaperObjectVisual
ICCV 2025-Selective Contrastive Learning for Weakly Supervised Affordance GroundingPaperObjectVisual
ICCV 2025OVA-FieldsOVA-Fields: Weakly Supervised Open-Vocabulary Affordance Fields for Robot Operational Part DetectionPaperObjectSpatial · Generalizable
CVPR 2025FunGraph3DOpen-Vocabulary Functional 3D Scene Graphs for Real-World Indoor SpacesPaperSceneSpatial
ACM MM 2025RoboAffordRoboAfford: A Dataset and Benchmark for Enhancing Object and Spatial Affordance Learning in Robot ManipulationPaperObject/SceneVisual · Spatial
ECCV 2026AffogatoAffogato: Learning Open-Vocabulary Affordance Grounding with Automated Data Generation at ScalePaperObjectVisual · Spatial · Generalizable
CVPR 2026AffostructionAffostruction: 3D Affordance Grounding with Generative ReconstructionPaperObjectSpatial
arXiv 2026PanoAffordanceNetPanoAffordanceNet: Towards Holistic Affordance Grounding in 360° Indoor EnvironmentsPaperSceneVisual
ECCV 2026PAPPanoramic Affordance PredictionPaperSceneVisual
CVPR 2026AffordMatcherAffordMatcher: Affordance Learning in 3D Scenes from Visual SignifiersPaperSceneSpatial
ECCV 2026DAGDiffusion Models are Open-World Affordance Learners: Leveraging Generative Priors for 3D Affordance LearningPaperObjectSpatial · Generalizable
CVPR 2026FunRECFunREC: Reconstructing Functional 3D Scenes from Egocentric Interaction VideosPaperSceneInteraction-driven · Spatial
arXiv 2026AFUNAFUN: Towards an Affordance Foundation Model for Functionality UnderstandingPaperObjectVisual · Spatial · Generalizable

🧠 Affordance Reasoning

Venue/DateNameTitlePaperReasoning Subcategory
ECCV 2014-Reasoning about Object Affordances in a Knowledge Base RepresentationPaperRelation-Based Affordance Reasoning
TCDS 2016-Bootstrapping Relational Affordances of Object Pairs Using TransferPaperRelation-Based Affordance Reasoning
CVPR 2018-Learning to Act Properly: Predicting and Explaining Affordances From ImagesPaperRelation-Based Affordance Reasoning
RA-L/IROS 2019-Learning Grasp Affordance Reasoning through Semantic RelationsPaperRelation-Based Affordance Reasoning
ICRA 2020-Learning Object Placements For Relational Instructions by Hallucinating Scene RepresentationsPaperRelation-Based Affordance Reasoning
arXiv 2021AR-NetRelationship Oriented Affordance Learning through Manipulation Graph ConstructionPaperRelation-Based Affordance Reasoning
ICDH 2022VAR-NetA Visual Affordance Reasoning Network Based on Graph AttentionPaperRelation-Based Affordance Reasoning
CoRL 2023SayCanDo As I Can, Not As I Say: Grounding Language in Robotic AffordancesPaperAgentic Affordance Reasoning
NeurIPS 2023Grounded DecodingGrounded Decoding: Guiding Text Generation with Grounded Models for Embodied AgentsPaperAgentic Affordance Reasoning
CVPRW 2024AffordanceLLMAffordanceLLM: Grounding Affordance from Vision Language ModelsPaperLanguage-Centric Affordance Reasoning
RA-L 2024NaturalVLMNaturalVLM: Leveraging Fine-grained Natural Language for Affordance-Guided Visual ManipulationPaperLanguage-Centric Affordance Reasoning
ICRAW 2024OVAL-PromptOVAL-Prompt: Open-Vocabulary Affordance Localization for Robot Manipulation through LLM Affordance-GroundingPaperLanguage-Centric Affordance Reasoning
ICTAI 2024WorldAffordWorldAfford: Affordance Grounding based on Natural Language InstructionsPaperLanguage-Centric Affordance Reasoning
CoRL 2024RoboPointRoboPoint: A Vision-Language Model for Spatial Affordance Prediction for RoboticsPaperLanguage-Centric Affordance Reasoning
IROS 2024ManipVQAManipVQA: Injecting Robotic Affordance and Physically Grounded Information into Multi-Modal Large Language ModelsPaperLanguage-Centric Affordance Reasoning
IROS 2025PAVLMPAVLM: Advancing Point Cloud based Affordance Understanding Via Vision-Language ModelPaperLanguage-Centric Affordance Reasoning
CVPR 2025GREATGREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance GroundingPaperLanguage-Centric Affordance Reasoning
CVPR 2025SeqAffordSeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language ModelPaperLanguage-Centric Affordance Reasoning
ICLR 20253D-AffordanceLLM3D-AffordanceLLM: Harnessing Large Language Models for Open-Vocabulary Affordance Detection in 3D WorldsPaperLanguage-Centric Affordance Reasoning
arXiv 2025Afford-XAfford-X: Generalizable and Slim Affordance Reasoning for Task-oriented ManipulationPaperLanguage-Centric Affordance Reasoning
CVPR 2025LMAffordance3DGrounding 3D Object Affordance with Language Instructions, Visual Observations and InteractionsPaperLanguage-Centric Affordance Reasoning
ACM MM 20253DAffordSplat3DAffordSplat: Efficient Affordance Reasoning with 3D GaussiansPaperLanguage-Centric Affordance Reasoning
ICRA 2025UniAffUniAff: A Unified Representation of Affordances for Tool Usage and Articulation with Vision-Language ModelsPaperLanguage-Centric Affordance Reasoning
ACL 2025InstructPartInstructPart: Task-Oriented Part Segmentation with Instruction ReasoningPaperLanguage-Centric Affordance Reasoning
CVPR 2025Fun3DUFunctionality Understanding and Segmentation in 3D ScenesPaperLanguage-Centric Affordance Reasoning
ICCV 2025RAGNetRAGNet: Large-scale Reasoning-based Affordance Segmentation Benchmark towards General GraspingPaperLanguage-Centric Affordance Reasoning
arXiv 2025SeqAffordSplatSeqAffordSplat: Scene-level Sequential Affordance Reasoning on 3D Gaussian SplattingPaperLanguage-Centric Affordance Reasoning
arXiv 2025ASPAgentic Scene Policies: Unifying Space, Semantics, and Affordances for Robot ActionPaperAgentic Affordance Reasoning
ACM MM 2025Aff3DFuncOpen-Vocabulary 3D Affordance Understanding via Functional Text Enhancement and Multilevel Representation AlignmentPaperLanguage-Centric Affordance Reasoning
IROS 2025AffordGraspAffordGrasp: In-Context Affordance Reasoning for Open-Vocabulary Task-Oriented Grasping in ClutterPaperLanguage-Centric Affordance Reasoning
NeurIPS 2025AffordBotAffordBot: 3D Fine-grained Embodied Reasoning via Multimodal Large Language ModelsPaperLanguage-Centric Affordance Reasoning
IROSW 2025RoboAfford++RoboAfford++: A Generative AI-Enhanced Dataset for Multimodal Affordance Learning in Robotic Manipulation and NavigationPaperLanguage-Centric Affordance Reasoning
NeurIPS 2025ViSPLAViSPLA: Visual Iterative Self-Prompting for Language-Guided 3D Affordance LearningPaperLanguage-Centric Affordance Reasoning
ICASSP 2026A4BenchAffordance Benchmark for MLLMsPaperLanguage-Centric Affordance Reasoning
AAAI 2026Affordance-R1Affordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language ModelPaperLanguage-Centric Affordance Reasoning
AAAI 2026TASATask-Aware 3D Affordance Segmentation via 2D Guidance and Geometric RefinementPaperLanguage-Centric Affordance Reasoning
ECCV 2026A4-AgentA4-Agent: An Agentic Framework for Zero-Shot Affordance ReasoningPaperAgentic Affordance Reasoning
arXiv 2026TRACERTRACER: Texture-Robust Affordance Chain-of-Thought for Deformable-Object RefinementPaperLanguage-Centric Affordance Reasoning
arXiv 2026AffordanceGrasp-R1AffordanceGrasp-R1: Leveraging Reasoning-Based Affordance Segmentation with Reinforcement Learning for Robotic GraspingPaperLanguage-Centric Affordance Reasoning
arXiv 2026VideoAffordVideoAfford: Grounding 3D Affordance from Human-Object-Interaction Videos via Multimodal Large Language ModelPaperLanguage-Centric Affordance Reasoning
arXiv 2026SceneTeractSceneTeract: Agentic Functional Affordances and VLM Grounding in 3D ScenesPaperAgentic Affordance Reasoning
arXiv 2026-Part-Aware Open-Vocabulary 3D Affordance Grounding via Prototypical Semantic and Geometric AlignmentPaperLanguage-Centric Affordance Reasoning
arXiv 2026A3RA3R: Agentic Affordance Reasoning via Cross-Dimensional Evidence in 3D Gaussian ScenesPaperAgentic Affordance Reasoning
ECCV 2026TokAGToken-Based Affordance Grounding with Large Vision-Language ModelsPaperLanguage-Centric Affordance Reasoning
arXiv 2026CompassADCompassAD: Intent-Driven 3D Affordance Grounding in Functionally Competing ObjectsPaperLanguage-Centric Affordance Reasoning
arXiv 2026A-HarnessAffordance Agent Harness: Verification-Gated Skill OrchestrationPaperAgentic Affordance Reasoning
arXiv 2026AffordMEMGrounding by Remembering: Cross-Scene and In-Scene Memory for 3D Functional AffordancesPaperAgentic Affordance Reasoning

🤖 Affordance-Guided Action

Venue/DateNameTitlePaperArchitectureAffordance UsageAction Subcategory
IEEE TCDS 2016-Training Agents With Interactive Reinforcement Learning and Contextual AffordancesPaperE2ErewardIntegrated policy · Optimization signal
ICRA 2020ASPNLearning Affordance Space in Physical World for Vision-based Robotic Object ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2021GRAFFLearning Dexterous Grasping with Object-Centric Visual AffordancesPaperE2ErewardIntegrated policy · Optimization signal
ICRA 2021VALWhat Can I Do Here? Learning New Skills by Imagining Visual AffordancesPaperE2ErewardIntegrated policy · Optimization signal
ICRA 2021Contact-GraspNetContact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered ScenesPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2021Where2ActWhere2Act: From Pixels to Actions for Articulated 3D ObjectsPaperHierarchical modelexplicitHierarchical affordance-to-action
ECCV 2022AdaAffordAdaAfford: Learning to Adapt Manipulation Affordance for 3D Articulated Objects via Few-shot InteractionsPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2022VAPOAffordance Learning from Play for Sample-Efficient Policy LearningPaperE2ErewardIntegrated policy · Optimization signal
IROS 2022-Learning 6-DoF Task-oriented Grasp Detection via Implicit Estimation and Visual AffordancePaperHierarchical modelexplicitHierarchical affordance-to-action
ICLR 2022VAT-MartVAT-Mart: Learning Visual Action Trajectory Proposals for Manipulating 3D Articulated ObjectsPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2023HULC++Grounding Language with Visual Affordances over Unstructured DataPaperE2EimplicitIntegrated policy · Implicit representation
ICLR 2023DualAffordDualAfford: Learning Collaborative Visual Affordance for Dual-gripper Object ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2023RLAffordRLAfford: End-to-End Affordance Learning for Robotic ManipulationPaperE2EimplicitIntegrated policy · Implicit representation
ICRA 2023-Visual Affordance Prediction for Guiding Robot ExplorationPaperE2ErewardIntegrated policy · Optimization signal
CoRL 2023ACE-NBVAffordance-Driven Next-Best-View Planning for Robotic GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
NeurIPS 2023-Learning Environment-aware Affordance for 3D Articulated Object Manipulation under OcclusionsPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2023LERF-TOGOLanguage Embedded Radiance Fields for Zero-Shot Task-Oriented GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2023VoxPoserVoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language ModelsPaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2023VRBAffordances from Human Videos as a Versatile Representation for RoboticsPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2023DefoAffordLearning Foresightful Dense Visual Affordance for Deformable Object ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IEEE T-RO 2024PRIMPPRIMP: Probabilistically-Informed Motion Primitives for Efficient Affordance Learning from DemonstrationPaperMotion primitivesexplicitHierarchical affordance-to-action · Primitive-based
arXiv 2024AnyPartOpen-Vocabulary Part-Based GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2024IDAInformation-driven Affordance Discovery for Efficient Robotic ManipulationPaperHierarchical modelrewardHierarchical affordance-to-action · Planner-based
arXiv 2024-Affordance-based Robot Manipulation with Flow MatchingPaperE2EexplicitIntegrated policy · Explicit input
CoRL 2024ReKepReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ECCV 2024Robo-ABCRobo-ABC: Affordance Generalization Beyond Categories via Semantic Correspondence for Robot ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IROS 2024PreAffordPreAfford: Universal Affordance-Based Pre-Grasping for Diverse Objects and EnvironmentsPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2024RAMRAM: Retrieval-Based Affordance Transfer for Generalizable Zero-Shot Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2025GATLearning Precise Affordances from Egocentric Videos for Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2025RT-AffordanceRT-Affordance: Affordances are Versatile Intermediate Representations for Robot ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IROS 2025ManipGPTManipGPT: Is Affordance Segmentation by Large Vision Models Enough for Articulated Object Manipulation?PaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2025AffordDexGraspAffordDexGrasp: Open-set Language-guided Dexterous Grasp with Generalizable-Instructive AffordancePaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2025GarmentPileGarmentPile: Point-Level Visual Affordance Guided Retrieval and Adaptation for Cluttered Garments ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2025A0A0: An Affordance-Aware Hierarchical Model for General Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IEEE CBS 2025DORADORA: Object Affordance-Guided Reinforcement Learning for Dexterous Robotic ManipulationPaperE2ErewardIntegrated policy · Optimization signal
CoRL 2025GLOVER++GLOVER++: Unleashing the Potential of Affordance Learning from Human Behaviors for Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2025AffordDPAffordDP: Generalizable Diffusion Policy with Transferable AffordancePaperE2EexplicitIntegrated policy · Explicit input
ICML 2025BiAssembleBiAssemble: Learning Collaborative Affordance for Bimanual Geometric AssemblyPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2025O3AffordO3Afford: One-Shot 3D Object-to-Object Affordance Grounding for Generalizable Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IEEE/ASME Transactions on Mechatronics 2025DARTDART: A Novel Task-Driven Diffusion-Based Policy with Affordance Learning for Generalizable Manipulation of Articulated ObjectsPaperE2EexplicitIntegrated policy · Explicit input
ICCV 2025CoA-VLACoA-VLA: Improving Vision-Language-Action Models via Visual-Textual Chain-of-AffordancePaperE2EexplicitIntegrated policy · Explicit input
ICCV 20252HandedAfforder2HandedAfforder: Learning Precise Actionable Bimanual Affordances from Human VideosPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2025OVAL-GraspOVAL-Grasp: Open-Vocabulary Affordance Localization for Task Oriented GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
RA-L 2025ScaleADFGScaleADFG: Affordance-based Dexterous Functional Grasping via Scalable DatasetPaperHierarchical modelexplicitHierarchical affordance-to-action
AAAI 2026AffordDexTowards Affordance-Aware Robotic Dexterous Grasping with Human-like PriorsPaperE2ErewardIntegrated policy · Optimization signal
CVPR 2026AFIAffordance Field Intervention: Enabling VLAs to Escape Memory Traps in Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
AAAI 2026A3DA3D: Adaptive Affordance Assembly with Dual-Arm ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IROS 2026FSAGFSAG: Enhancing Human-to-Dexterous-Hand Finger-Specific Affordance Grounding via Diffusion ModelsPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026CorDexGenerate, Transfer, Adapt: Learning Functional Dexterous Grasping from a Single Human DemonstrationPaperHierarchical modelimplicitHierarchical affordance-to-action · Retrieval-based
IEEE Access 2026AffoRo-GSOne-Shot 3-D Affordance Learning for Multi-Stage Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2026BiPreManipBiPreManip: Learning Affordance-Based Bimanual Preparatory Manipulation through Anticipatory CollaborationPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026RAAPRAAP: Retrieval-Augmented Affordance Prediction with Cross-Image Action AlignmentPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026BridgeACTBridgeACT: Bridging Human Demonstrations to Robot Actions via Unified Tool-Target AffordancesPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026Afford-VLAAfford-VLA: Action-Aligned Visual Planning via Internalized AffordancePaperE2EimplicitIntegrated policy · Implicit representation
arXiv 2026AffordVLAAffordVLA: Injecting Affordance Representations into Vision-Language-Action Models via Implicit Feature AlignmentPaperE2EimplicitIntegrated policy · Implicit representation
CVPR 2026PALMPALM: Progress-Aware Policy Learning via Affordance Reasoning for Long-Horizon Robotic ManipulationPaperE2EimplicitIntegrated policy · Implicit representation
CVPR 2026AffordGenAffordGen: Generating Diverse Demonstrations for Generalizable Object Manipulation with Affordance CorrespondencePaperE2EimplicitIntegrated policy · Implicit representation
arXiv 2026AffordanceVLAAffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware UnderstandingPaperE2EimplicitIntegrated policy · Implicit representation
arXiv 2026Affordance2ActionAffordance2Action: Task-Conditioned Scene-level Affordance Grounding for Real-Time ManipulationPaperE2Eexplicit/implicitIntegrated policy · Explicit/implicit representation
arXiv 2026GROW2GROW2: Grounding Which and Where for Robot Tool UsePaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026AffordSimAffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action · Simulation benchmark

📊 Affordance Datasets / Benchmarks

Survey groups: object-centric · scene-level · interaction-driven · language- and reasoning-oriented · action-oriented.

Venue/DateNameTitlePaperData Modality
ICRA 2015UMD RGB-D Part Affordance DatasetAffordance Detection of Tool Parts from Geometric FeaturesPaperimage
ECCV 2016NYUv2 affordance annotationsA Multi-scale CNN for Affordance Segmentation in RGB ImagesPaperimage
IROS 2017IIT-AFFObject-Based Affordances Detection with Convolutional Neural Networks and Dense Conditional Random FieldsPaperimage
CVPR 2018ADE-AffordanceLearning to Act Properly: Predicting and Explaining Affordances From ImagesPaperimage
CVPR 2018OPRADemo2Vec: Reasoning Object Affordances from Online VideosPapervideo-image
ECCV 2018EPIC-KITCHENSScaling egocentric vision: The epic-kitchens datasetPapervideo
CVPR 2020GraspNet-1BillionGraspnet-1billion: A large-scale benchmark for general object graspingPaperRGB-D/3D grasp
CVPR 20213D AffordanceNet3D AffordanceNet: A Benchmark for Visual Object Affordance UnderstandingPaper3D
CVPR 2022AGD20KLearning Affordance Grounding from Exocentric ImagesPaperimage-image
CVPR 2022Ego4DEgo4d: Around the world in 3,000 hours of egocentric videoPapervideo
ECCVW 2022PartAffordPartAfford: Part-level Affordance Discovery from 3D ObjectsPaper3D
ICCV 20233DOIUnderstanding 3D Object Interaction from a Single ImagePaperimage/3D interaction
ICCV 2023AffordPoseAffordPose: A Large-Scale Dataset of Hand-Object Interactions with Affordance-Driven Hand PosePaper3D
ICCV 2023EPIC-AffMulti-label Affordance Mapping from Egocentric VisionPapervideo-3D
ICCV 2023PIADGrounding 3D Object Affordance from 2D Interactions in ImagesPaperimage-3D
IROS 2023HANDALHANDAL: A Dataset of Real-World Manipulable Object Categories with Pose Annotations, Affordances, and ReconstructionsPaperRGB-D/6-DoF pose/3D mesh
WACV 2023EPIC-KITCHENS AffordanceFine-Grained Affordance Annotation for Egocentric Hand-Object Interaction VideosPapervideo/image
CVPR 20243DIRLEMON: Learning 3D Human-Object Interaction Relation from 2D ImagesPaperimage-3D
CVPR 2024AffordQLASO: Language-guided Affordance Segmentation on 3D ObjectPaper3D
CVPR 2024SceneFun3DSceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D ScenesPaper3D
ACL 2025InstructPartInstructPart: Task-Oriented Part Segmentation with Instruction ReasoningPaperimage-language
ACM MM 20253DAffordSplat3DAffordSplat: Efficient Affordance Reasoning with 3D GaussiansPaper3D
ACM MM 2025RoboAffordRoboAfford: A Dataset and Benchmark for Enhancing Object and Spatial Affordance Learning in Robot ManipulationPaperimage/language
arXiv 2025A4BenchAffordance Benchmark for MLLMsPaperimage
arXiv 2025Affogato-750KAffogato: Learning Open-Vocabulary Affordance Grounding with Automated Data Generation at ScalePaperlanguage-3D
arXiv 2025HOVA-500KGLOVER++: Unleashing the Potential of Affordance Learning from Human Behaviors for Robotic ManipulationPaperimage
arXiv 2025SeqAffordSplatSeqAffordSplat: Scene-level Sequential Affordance Reasoning on 3D Gaussian SplattingPaperlanguage/3D Gaussian
CVPR 2025AGPILGrounding 3D Object Affordance with Language Instructions, Visual Observations and InteractionsPaperimage-3D
CVPR 2025FunGraph3DOpen-Vocabulary Functional 3D Scene Graphs for Real-World Indoor SpacesPaperRGB-D/3D scene graph
CVPR 2025PIADv2GREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance GroundingPaperimage-language-3D
CVPR 2025SeqAffordSeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language ModelPaperlanguage-3D
ICCV 20252HandedAffordance2HandedAfforder: Learning Precise Actionable Bimanual Affordances from Human VideosPapervideo/text/3D
ICCV 2025RAGNetRAGNet: Large-scale Reasoning-based Affordance Segmentation Benchmark towards General GraspingPaperimage
IROS 2025AGDDO15One-Shot Affordance Grounding of Deformable Objects in Egocentric Organizing ScenesPaperimage
IROS Workshop 2025RoboAfford++RoboAfford++: A Generative AI-Enhanced Dataset for Multimodal Affordance Learning in Robotic Manipulation and NavigationPaperimage/language
RA-L 2025ScaleADFGScaleADFG: Affordance-based Dexterous Functional Grasping via Scalable DatasetPaper3D/grasp
AAAI 2026ReasonAffAffordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language ModelPaperimage/language
arXiv 2026360-AGDPanoAffordanceNet: Towards Holistic Affordance Grounding in 360° Indoor EnvironmentsPaperpanoramic image/text
arXiv 2026A2A-BenchAffordance2Action: Task-Conditioned Scene-level Affordance Grounding for Real-Time ManipulationPaperimage
arXiv 2026AffordSimAffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic ManipulationPapersimulation/3D/robot action
arXiv 2026CompassADCompassAD: Intent-Driven 3D Affordance Grounding in Functionally Competing ObjectsPaperlanguage-3D
arXiv 2026VIDAVideoAfford: Grounding 3D Affordance from Human-Object-Interaction Videos via Multimodal Large Language ModelPapervideo-3D
CVPR 2026AffordBridgeAffordMatcher: Affordance Learning in 3D Scenes from Visual SignifiersPaperimage-3D
ECCV 2026PAP-12KPanoramic Affordance PredictionPaperpanoramic image/text

📚 Related Surveys

Venue/DateNameTitlePaperFocus
IEEE TCDS 2016-Affordance Research in Developmental Robotics: A SurveyPaperDevelopmental robotics and robot acquisition of affordance knowledge
IEEE TCDS 2018-Affordances in Psychology, Neuroscience, and Robotics: A SurveyPaperCross-disciplinary foundations spanning psychology, neuroscience, and robotics
IJCAI 2021-Building Affordance Relations for Robotic Agents - A ReviewPaperRelational affordance representations for robotic agents
ACM Computing Surveys 2021-Visual Affordance and Function Understanding: A SurveyPaperVisual affordance recognition, segmentation, and function understanding
IEEE TBD 2023-A Survey of Visual Affordance Recognition Based on Deep LearningPaperDeep-learning methods for visual affordance recognition
IEEE TCDS 2023-Recent Advances of Deep Robotic Affordance Learning: A Reinforcement Learning PerspectivePaperRobotic affordance learning from a reinforcement-learning perspective
arXiv 2025-Visual Affordance Prediction: Survey and ReproducibilityPaperVisual affordance prediction with an emphasis on reproducibility

🎉 Contributing

⭐ Help us grow this repository! If you know any valuable works we’ve missed, don’t hesitate to contribute — every suggestion makes a difference!

We welcome and appreciate all contributions! Here’s how you can help:

  • 📄 Add or Update a Paper
    Contribute by adding a new paper or improving details of an existing one. Please consider the most appropriate category for the work.

  • ✍️ Use Consistent Formatting
    Follow the format of the existing entries to maintain clarity and consistency across the list.

  • 🔗 Include Abstract Link
    If the paper is from arXiv, use the /abs/ link format for the abstract (e.g., https://arxiv.org/abs/xxxx.xxxxx).

  • 💡 Explain Your Edit (Optional but Helpful)
    A short note on why you think the paper deserves to be added or updated is appreciated and helps maintainers process your PR faster.

✅ Don't worry about getting everything perfect!
Minor mistakes are totally fine — we’ll help fix them. What matters most is your contribution. Let's highlight your awesome work together!

🌟 Acknowledgment

Thanks for the wonderful researchers focusing on affordance learning and embodied AI

📄 License

This project is licensed under the MIT License.

Contributors

hq-King

64 commits

Zhenhao-Zhang

2 commits

ha8siouhdi

1 commits

XubeiPan666

1 commits

Similar repos

YanjieZe/Paper-List

knowledge

603

BaiShuanghao/Awesome-Robotics-Manipulation

A comprehensive list of papers about Robot Manipulation, including papers, codes, and related…

1,151

lhc1224/Cross-View-AG

Official PyTorch Implementation of Learning Affordance Grounding from Exocentric Images, CVPR 2022

77

robotics-survey/Awesome-Robotics-Foundation-Models

1,411

Sid2697/awesome-egocentric-vision

A curated list of egocentric (first-person) vision and related area resources

348

yueen-ma/Awesome-VLA

671

DoranLyong/oh-my-world-intelligence

From 3D/4D Scene Understanding to Semantic-metric World Models

0

showlab/Awesome-Multimodal-Embodied-Agent

Survey on Multimodal Embodied Agents: From Computer-Use to Robot-Use

70

hq-King/Awesome-Affordance-Learning

229

68 commits

updated Sep 10, 2026

See the code

README

From Passive Perception to Active Interaction: A Survey of
Affordance Learning for Embodied AI

Awesome License Last Commit GitHub Stars

Gen Li1,#,‡, Hanqing Wang2,#, Jingliang Li1,#, Yifan Han3,#, Jindou Jia1, Tao Lin3, Yuanzhe Liu4, Yutong Wang5, Bo Zhao3, Fangqiang Ding2, Anh Nguyen6, Laura Sevilla-Lara7, Huazhe Xu8, Gregory S. Chirikjian9,
Marc Pollefeys10, Oier Mees10, Hui Xiong2,†, Jianfei Yang1,†

1Nanyang Technological University    2HKUST(GZ)    3SJTU 4UIUC    5The University of Sydney   
6University of Liverpool    7The University of Edinburgh 8MBZUAI    9Tsinghua University    10ETH Zurich

#Equal Contribution    ‡Project Lead    †Corresponding Author

A curated collection of papers on affordance learning for embodied AI.

🧭 Exploring Embodied AI and Embodied perception? We hope this collection proves useful in your journey. If you'd like to support the project, feel free to ⭐️ the repo and share it with your peers. Contributions are warmly welcome!

📖 Contents

🔥 News

📢 This list is actively maintained, and community contributions are always appreciated!
Feel free to open a pull request if you find any relevant papers.

  • [2025-05] 🎉 This repository was launched to curate a comprehensive list of affordance-learning research.

🌟 Introduction

This repository accompanies the survey From Passive Perception to Active Interaction: A Survey of Affordance Learning for Embodied AI and maintains a curated collection of papers, datasets, and benchmarks.

As robots and embodied agents move into real-world applications, they must understand not only what objects are, but also where, why, and how they can interact with them. Following the survey, the list is organized around three complementary questions:

  • Affordance Perception: Where or which region affords an interaction?
  • Affordance Reasoning: Which affordance is relevant under the current task, environment, and constraints?
  • Affordance-Guided Action: How can an affordance be converted into an executable action?

A paper may be cross-referenced when the survey discusses it in more than one role (for example, both perception and reasoning). A separate section collects datasets and benchmarks. Venue labels use the formally published version whenever one is available; otherwise the first public preprint is marked arXiv.

🧭 Taxonomy

The taxonomy follows the survey's functional pipeline rather than only model architecture. It expands each primary category into second- and third-level categories. Cross-listing is intentional when a method contributes to multiple stages or perspectives.

Primary CategorySecondary CategoryThird-Level CategoriesMain Distinction
Affordance PerceptionVisual affordance perceptionObject-centric grounding; scene-level grounding; weakly supervised perceptionGrounds masks, heatmaps, or keypoints on the 2D image plane.
Affordance PerceptionSpatial affordance perception3D object grounding; 3D scene groundingGrounds functional regions or geometric structures directly in metric 3D space.
Affordance PerceptionInteraction-driven affordance perceptionDemonstration/HOI video; HOI image; interaction-conditioned 3D; interaction-grounded scene perceptionDerives supervision or context from observed human-object interactions.
Affordance PerceptionGeneralizable affordance perceptionExample-based transfer; open-set groundingTransfers to novel objects, labels, queries, or interaction contexts.
Affordance ReasoningRelation-based reasoningProbabilistic/semantic relations; object-pair and scene context; manipulation graphsInfers affordances through explicit relations among objects, actions, agents, and scene context.
Affordance ReasoningLanguage-centric reasoningGrounded skill selection; integrated LLM/MLLM prediction; modular semantic-to-spatial grounding; sequential reasoningUses language to interpret intent and select or ground task-relevant affordances.
Affordance ReasoningAgentic reasoningPredefined workflows; adaptive workflowsUses planning, memory, verification, or iterative tool/model calls over multiple steps.
Affordance-Guided ActionHierarchical affordance-to-actionPrimitive-based execution; retrieval-based transfer; planner-based executionPredicts affordances first and then converts them into actions through a separate execution module.
Affordance-Guided ActionAffordance-integrated policy learningAffordance as explicit input; implicit representation; optimization signalIntegrates affordance information directly into policy representation, input, or optimization.

📄 Paper List

The Venue/Date column prioritizes the formal venue and publication year. Papers without a confirmed venue are labeled arXiv. The Name column records the method or system name explicitly introduced by the authors, including names stated only in the abstract or main text rather than in the title; - means that no explicit method name has been verified and no acronym is inferred from the title.

👁️ Affordance Perception

Venue/DateNameTitlePaperScopePerception Subcategory
IJRR 2013-Learning Human Activities and Object Affordances from RGB-D VideosPaperSceneInteraction-driven
ICRA 2015UMDAffordance Detection of Tool Parts from Geometric FeaturesPaperObjectVisual · Spatial
ECCV 2016-A Multi-scale CNN for Affordance Segmentation in RGB ImagesPaperSceneVisual
IROS 2016-Detecting Object Affordances with Convolutional Neural NetworksPaperObjectVisual
CVPR 2017-Weakly Supervised Affordance DetectionPaperObjectVisual
IROS 2017-Object-Based Affordances Detection with Convolutional Neural Networks and Dense Conditional Random FieldsPaperObjectVisual
ICCVW 2017-Adaptive Binarization for Weakly Supervised Affordance SegmentationPaperObjectVisual
ICCVW 2017-Learning to Segment AffordancesPaperObject/SceneVisual
Humanoids 2017-Affordance Detection for Task-Specific Grasping Using Deep LearningPaperObjectVisual
ICRA 2018AffordanceNetAffordanceNet: An End-to-End Deep Learning Approach for Object Affordance DetectionPaperObjectVisual
CVPR 2018Demo2VecDemo2Vec: Reasoning Object Affordances from Online VideosPaperObjectVisual · Interaction-driven
ECCV 2018Gaze + ActionIn the Eye of Beholder: Joint Learning of Gaze and Actions in First Person VideoPaperSceneInteraction-driven
Autonomous Robots 2019-Towards Affordance Detection for Robot Manipulation Using Affordance for Parts and Parts for AffordancePaperObjectVisual
RAS 2019-Context-Based Affordance Segmentation from 2D Images for Robot ActionsPaperObjectVisual
RA-L with ICRA 2019AffordanceLearnLearning Affordance Segmentation for Real-world Robotic Manipulation via Synthetic ImagesPaperObjectVisual
arXiv 2019AffContextRecognizing Object Affordances to Support Scene Reasoning for Manipulation TasksPaperObjectVisual
ICCV 2019Interaction HotspotsGrounded Human-Object Interaction Hotspots From VideoPaperObjectVisual · Interaction-driven
CVPR 2020Ego-TopoEgo-Topo: Environment Affordances from Egocentric VideoPaperSceneInteraction-driven
NeurIPS 2020IntExpLearning Affordance Landscapes for Interaction Exploration in 3D EnvironmentsPaperSceneSpatial
Neural Comput. Appl. 2020.09-Object Affordance Detection with Relationship-Aware NetworkPaperObjectVisual
RA-L with ICRA2021AffKpAn Affordance Keypoint Detection Network for Robot ManipulationPaperObjectVisual
ACM ICDAR 2021ST-HOIST-HOI: A Spatial-Temporal Baseline for Human-Object Interaction Detection in VideosPaperSceneInteraction-driven
CVPR 20213D AffordanceNet3D AffordanceNet: A Benchmark for Visual Object Affordance UnderstandingPaperObjectSpatial
CoRL 2021O2O-AffordO2O-Afford: Annotation-Free Large-Scale Object-Object Affordance LearningPaperObject-pairSpatial
Neurocomputing 2021-Visual Affordance Detection Using an Efficient Attention Convolutional Neural NetworkPaperObjectVisual
IJCAI 2021OS-ADOne-Shot Affordance DetectionPaperObjectVisual · Generalizable
Humanoids 2022ACFManipulation-Oriented Object Perception in Clutter through Affordance Coordinate FramesPaperObjectSpatial
ECCVW 2022PartAffordPartAfford: Part-level Affordance Discovery from 3D ObjectsPaperObjectSpatial
CVPR 2022OCTJoint Hand Motion and Interaction Hotspots Prediction from Egocentric VideosPaperObjectVisual · Interaction-driven
CVPR 2022Cross-view-AGLearning Affordance Grounding from Exocentric ImagesPaperObjectVisual · Interaction-driven
CoRL 2022AffCorrsOne-Shot Transfer of Affordance Regions? AffCorrs!PaperObjectVisual · Generalizable
Neural Comput. Appl. 2022BPNObject Affordance Detection with Boundary-Preserving Network for Robotic Manipulation TasksPaperObjectVisual
IJCV 2022OSAD-NetOne-Shot Object Affordance Detection in the WildPaperObjectVisual · Generalizable
WACV 2023-Fine-Grained Affordance Annotation for Egocentric Hand-Object Interaction VideosPaperObjectInteraction-driven
arXiv 2023STRAPSTRAP: Structured Object Affordance Segmentation with Point SupervisionPaperObjectVisual
CVPR 2023AfformerAffordance Grounding From Demonstration Video To Target ImagePaperObjectInteraction-driven
CVPR 2023Leverage Interactive Affinity for Affordance LearningPaperObjectVisual · Interaction-driven
CVPR 2023LOCATELOCATE: Localize and Transfer Object Parts for Weakly Supervised Affordance GroundingPaperObjectVisual · Interaction-driven
NeurIPS 2023Where2ExploreWhere2Explore: Few-shot Affordance Learning for Unseen Novel Categories of Articulated ObjectsPaperObjectSpatial · Generalizable
ICIP 2023-A Large Scale Multi-View RGBD Visual Affordance Learning DatasetPaperObjectVisual
IROS 2023VATHierarchical Transformer for Visual Affordance Understanding using a Large-scale DatasetPaperObjectVisual
ICCV 2023AffordPoseAffordPose: A Large-Scale Dataset of Hand-Object Interactions with Affordance-Driven Hand PosePaperObjectSpatial · Interaction-driven
IROS 2023HANDALHANDAL: A Dataset of Real-World Manipulable Object Categories with Pose Annotations, Affordances, and ReconstructionsPaperObjectVisual · Spatial
ICCV 2023IAGNetGrounding 3D Object Affordance from 2D Interactions in ImagesPaperObjectSpatial · Interaction-driven
ICCV 2023-Understanding 3D Object Interaction from a Single ImagePaperSceneSpatial
ICCV 2023EPIC-AffMulti-label Affordance Mapping from Egocentric VisionPaperSceneSpatial · Interaction-driven
IROS 2023OpenADOpen-vocabulary Affordance Detection in 3D Point CloudsPaperObjectSpatial · Generalizable
ICRA 2024-Open-Vocabulary Affordance Detection using Knowledge Distillation and Text-Point CorrelationPaperObjectSpatial · Generalizable
CVPR 2024LEMONLEMON: Learning 3D Human-Object Interaction Relation from 2D ImagesPaperObjectSpatial · Interaction-driven
AAAI 2024WSMAWeakly Supervised Multimodal Affordance Grounding for Egocentric ImagesPaperObjectVisual · Interaction-driven
arXiv 2024-Text-driven Affordance Learning from Egocentric VisionPaperObjectInteraction-driven
ICRA 20243DAPNetLanguage-Conditioned Affordance-Pose Detection in 3D Point CloudsPaperObjectSpatial · Generalizable
CVPR 2024OOALOne-Shot Open Affordance Learning with Foundation ModelsPaperObjectVisual · Generalizable
CVPR 2024LASOLASO: Language-guided Affordance Segmentation on 3D ObjectPaperObjectSpatial · Generalizable
CVPRW 2024LGAfford-NetLGAfford-Net: A Local Geometry Aware Affordance Detection Network for 3D Point CloudsPaperObjectSpatial
CVPR 2024SceneFun3DSceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D ScenesPaperSceneSpatial
arXiv 2024Ego-SAGGrounding 3D Scene Affordance From Egocentric InteractionsPaperSceneInteraction-driven · Spatial
ECCV 2024INTRAINTRA: Interaction Relationship-aware Weakly Supervised Affordance GroundingPaperObjectVisual · Interaction-driven
ECCV 2024ComABeyond the Contact: Discovering Comprehensive Affordance for 3D Objects from Pre-trained 2D Diffusion ModelsPaperObjectSpatial
IROS 2024FAKP-Net3D Affordance Keypoint Detection for Robotic ManipulationPaperObjectSpatial
ICCV 2025GLANCEGLANCE: Intermediate Connectors and Geometric Priors for Language-Guided Affordance Segmentation on Unseen Object CategoriesPaperObjectSpatial · Generalizable
AAAI 2025MIFAGLearning 2D Invariant Affordance Knowledge for 3D Affordance GroundingPaperObjectSpatial · Interaction-driven
CVPR 2025GEALGEAL: Generalizable 3D Affordance Learning with Cross-Modal ConsistencyPaperObjectSpatial · Generalizable
AAAI 2025MaskPromptMaskPrompt: Open-Vocabulary Affordance Segmentation with Object Shape Mask PromptsPaperObjectVisual · Generalizable
arXiv 2025AffordanceSAMAffordanceSAM: Segment Anything Once More in Affordance GroundingPaperObjectVisual · Generalizable
CVPR 2025IAAOIAAO: Interactive Affordance Learning for Articulated Objects in 3D EnvironmentsPaperObjectSpatial
arXiv 2025-Interpretable Affordance Detection on 3D Point Clouds with Probabilistic PrototypesPaperObjectSpatial
arXiv 2025-Weakly-Supervised Affordance Grounding Guided by Part-Level Semantic PriorsPaperObjectVisual
ICRA 2025UADUAD: Unsupervised Affordance Distillation for Generalization in Robotic ManipulationPaperObjectVisual · Generalizable
CVPR 2025Reasoning MambaReasoning Mamba: Hypergraph-Guided Region Relation Calculating for Weakly Supervised Affordance GroundingPaperObjectVisual
IROS 2025OS-AGDOOne-Shot Affordance Grounding of Deformable Objects in Egocentric Organizing ScenesPaperObjectVisual · Generalizable
IROS 2025BiT-AlignResource-Efficient Affordance Grounding with Complementary Depth and Semantic PromptsPaperObjectVisual
ICCV 2025LoopTransClosed-Loop Transfer for Weakly-supervised Affordance GroundingPaperObjectVisual
ICCV 2025-Selective Contrastive Learning for Weakly Supervised Affordance GroundingPaperObjectVisual
ICCV 2025OVA-FieldsOVA-Fields: Weakly Supervised Open-Vocabulary Affordance Fields for Robot Operational Part DetectionPaperObjectSpatial · Generalizable
CVPR 2025FunGraph3DOpen-Vocabulary Functional 3D Scene Graphs for Real-World Indoor SpacesPaperSceneSpatial
ACM MM 2025RoboAffordRoboAfford: A Dataset and Benchmark for Enhancing Object and Spatial Affordance Learning in Robot ManipulationPaperObject/SceneVisual · Spatial
ECCV 2026AffogatoAffogato: Learning Open-Vocabulary Affordance Grounding with Automated Data Generation at ScalePaperObjectVisual · Spatial · Generalizable
CVPR 2026AffostructionAffostruction: 3D Affordance Grounding with Generative ReconstructionPaperObjectSpatial
arXiv 2026PanoAffordanceNetPanoAffordanceNet: Towards Holistic Affordance Grounding in 360° Indoor EnvironmentsPaperSceneVisual
ECCV 2026PAPPanoramic Affordance PredictionPaperSceneVisual
CVPR 2026AffordMatcherAffordMatcher: Affordance Learning in 3D Scenes from Visual SignifiersPaperSceneSpatial
ECCV 2026DAGDiffusion Models are Open-World Affordance Learners: Leveraging Generative Priors for 3D Affordance LearningPaperObjectSpatial · Generalizable
CVPR 2026FunRECFunREC: Reconstructing Functional 3D Scenes from Egocentric Interaction VideosPaperSceneInteraction-driven · Spatial
arXiv 2026AFUNAFUN: Towards an Affordance Foundation Model for Functionality UnderstandingPaperObjectVisual · Spatial · Generalizable

🧠 Affordance Reasoning

Venue/DateNameTitlePaperReasoning Subcategory
ECCV 2014-Reasoning about Object Affordances in a Knowledge Base RepresentationPaperRelation-Based Affordance Reasoning
TCDS 2016-Bootstrapping Relational Affordances of Object Pairs Using TransferPaperRelation-Based Affordance Reasoning
CVPR 2018-Learning to Act Properly: Predicting and Explaining Affordances From ImagesPaperRelation-Based Affordance Reasoning
RA-L/IROS 2019-Learning Grasp Affordance Reasoning through Semantic RelationsPaperRelation-Based Affordance Reasoning
ICRA 2020-Learning Object Placements For Relational Instructions by Hallucinating Scene RepresentationsPaperRelation-Based Affordance Reasoning
arXiv 2021AR-NetRelationship Oriented Affordance Learning through Manipulation Graph ConstructionPaperRelation-Based Affordance Reasoning
ICDH 2022VAR-NetA Visual Affordance Reasoning Network Based on Graph AttentionPaperRelation-Based Affordance Reasoning
CoRL 2023SayCanDo As I Can, Not As I Say: Grounding Language in Robotic AffordancesPaperAgentic Affordance Reasoning
NeurIPS 2023Grounded DecodingGrounded Decoding: Guiding Text Generation with Grounded Models for Embodied AgentsPaperAgentic Affordance Reasoning
CVPRW 2024AffordanceLLMAffordanceLLM: Grounding Affordance from Vision Language ModelsPaperLanguage-Centric Affordance Reasoning
RA-L 2024NaturalVLMNaturalVLM: Leveraging Fine-grained Natural Language for Affordance-Guided Visual ManipulationPaperLanguage-Centric Affordance Reasoning
ICRAW 2024OVAL-PromptOVAL-Prompt: Open-Vocabulary Affordance Localization for Robot Manipulation through LLM Affordance-GroundingPaperLanguage-Centric Affordance Reasoning
ICTAI 2024WorldAffordWorldAfford: Affordance Grounding based on Natural Language InstructionsPaperLanguage-Centric Affordance Reasoning
CoRL 2024RoboPointRoboPoint: A Vision-Language Model for Spatial Affordance Prediction for RoboticsPaperLanguage-Centric Affordance Reasoning
IROS 2024ManipVQAManipVQA: Injecting Robotic Affordance and Physically Grounded Information into Multi-Modal Large Language ModelsPaperLanguage-Centric Affordance Reasoning
IROS 2025PAVLMPAVLM: Advancing Point Cloud based Affordance Understanding Via Vision-Language ModelPaperLanguage-Centric Affordance Reasoning
CVPR 2025GREATGREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance GroundingPaperLanguage-Centric Affordance Reasoning
CVPR 2025SeqAffordSeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language ModelPaperLanguage-Centric Affordance Reasoning
ICLR 20253D-AffordanceLLM3D-AffordanceLLM: Harnessing Large Language Models for Open-Vocabulary Affordance Detection in 3D WorldsPaperLanguage-Centric Affordance Reasoning
arXiv 2025Afford-XAfford-X: Generalizable and Slim Affordance Reasoning for Task-oriented ManipulationPaperLanguage-Centric Affordance Reasoning
CVPR 2025LMAffordance3DGrounding 3D Object Affordance with Language Instructions, Visual Observations and InteractionsPaperLanguage-Centric Affordance Reasoning
ACM MM 20253DAffordSplat3DAffordSplat: Efficient Affordance Reasoning with 3D GaussiansPaperLanguage-Centric Affordance Reasoning
ICRA 2025UniAffUniAff: A Unified Representation of Affordances for Tool Usage and Articulation with Vision-Language ModelsPaperLanguage-Centric Affordance Reasoning
ACL 2025InstructPartInstructPart: Task-Oriented Part Segmentation with Instruction ReasoningPaperLanguage-Centric Affordance Reasoning
CVPR 2025Fun3DUFunctionality Understanding and Segmentation in 3D ScenesPaperLanguage-Centric Affordance Reasoning
ICCV 2025RAGNetRAGNet: Large-scale Reasoning-based Affordance Segmentation Benchmark towards General GraspingPaperLanguage-Centric Affordance Reasoning
arXiv 2025SeqAffordSplatSeqAffordSplat: Scene-level Sequential Affordance Reasoning on 3D Gaussian SplattingPaperLanguage-Centric Affordance Reasoning
arXiv 2025ASPAgentic Scene Policies: Unifying Space, Semantics, and Affordances for Robot ActionPaperAgentic Affordance Reasoning
ACM MM 2025Aff3DFuncOpen-Vocabulary 3D Affordance Understanding via Functional Text Enhancement and Multilevel Representation AlignmentPaperLanguage-Centric Affordance Reasoning
IROS 2025AffordGraspAffordGrasp: In-Context Affordance Reasoning for Open-Vocabulary Task-Oriented Grasping in ClutterPaperLanguage-Centric Affordance Reasoning
NeurIPS 2025AffordBotAffordBot: 3D Fine-grained Embodied Reasoning via Multimodal Large Language ModelsPaperLanguage-Centric Affordance Reasoning
IROSW 2025RoboAfford++RoboAfford++: A Generative AI-Enhanced Dataset for Multimodal Affordance Learning in Robotic Manipulation and NavigationPaperLanguage-Centric Affordance Reasoning
NeurIPS 2025ViSPLAViSPLA: Visual Iterative Self-Prompting for Language-Guided 3D Affordance LearningPaperLanguage-Centric Affordance Reasoning
ICASSP 2026A4BenchAffordance Benchmark for MLLMsPaperLanguage-Centric Affordance Reasoning
AAAI 2026Affordance-R1Affordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language ModelPaperLanguage-Centric Affordance Reasoning
AAAI 2026TASATask-Aware 3D Affordance Segmentation via 2D Guidance and Geometric RefinementPaperLanguage-Centric Affordance Reasoning
ECCV 2026A4-AgentA4-Agent: An Agentic Framework for Zero-Shot Affordance ReasoningPaperAgentic Affordance Reasoning
arXiv 2026TRACERTRACER: Texture-Robust Affordance Chain-of-Thought for Deformable-Object RefinementPaperLanguage-Centric Affordance Reasoning
arXiv 2026AffordanceGrasp-R1AffordanceGrasp-R1: Leveraging Reasoning-Based Affordance Segmentation with Reinforcement Learning for Robotic GraspingPaperLanguage-Centric Affordance Reasoning
arXiv 2026VideoAffordVideoAfford: Grounding 3D Affordance from Human-Object-Interaction Videos via Multimodal Large Language ModelPaperLanguage-Centric Affordance Reasoning
arXiv 2026SceneTeractSceneTeract: Agentic Functional Affordances and VLM Grounding in 3D ScenesPaperAgentic Affordance Reasoning
arXiv 2026-Part-Aware Open-Vocabulary 3D Affordance Grounding via Prototypical Semantic and Geometric AlignmentPaperLanguage-Centric Affordance Reasoning
arXiv 2026A3RA3R: Agentic Affordance Reasoning via Cross-Dimensional Evidence in 3D Gaussian ScenesPaperAgentic Affordance Reasoning
ECCV 2026TokAGToken-Based Affordance Grounding with Large Vision-Language ModelsPaperLanguage-Centric Affordance Reasoning
arXiv 2026CompassADCompassAD: Intent-Driven 3D Affordance Grounding in Functionally Competing ObjectsPaperLanguage-Centric Affordance Reasoning
arXiv 2026A-HarnessAffordance Agent Harness: Verification-Gated Skill OrchestrationPaperAgentic Affordance Reasoning
arXiv 2026AffordMEMGrounding by Remembering: Cross-Scene and In-Scene Memory for 3D Functional AffordancesPaperAgentic Affordance Reasoning

🤖 Affordance-Guided Action

Venue/DateNameTitlePaperArchitectureAffordance UsageAction Subcategory
IEEE TCDS 2016-Training Agents With Interactive Reinforcement Learning and Contextual AffordancesPaperE2ErewardIntegrated policy · Optimization signal
ICRA 2020ASPNLearning Affordance Space in Physical World for Vision-based Robotic Object ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2021GRAFFLearning Dexterous Grasping with Object-Centric Visual AffordancesPaperE2ErewardIntegrated policy · Optimization signal
ICRA 2021VALWhat Can I Do Here? Learning New Skills by Imagining Visual AffordancesPaperE2ErewardIntegrated policy · Optimization signal
ICRA 2021Contact-GraspNetContact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered ScenesPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2021Where2ActWhere2Act: From Pixels to Actions for Articulated 3D ObjectsPaperHierarchical modelexplicitHierarchical affordance-to-action
ECCV 2022AdaAffordAdaAfford: Learning to Adapt Manipulation Affordance for 3D Articulated Objects via Few-shot InteractionsPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2022VAPOAffordance Learning from Play for Sample-Efficient Policy LearningPaperE2ErewardIntegrated policy · Optimization signal
IROS 2022-Learning 6-DoF Task-oriented Grasp Detection via Implicit Estimation and Visual AffordancePaperHierarchical modelexplicitHierarchical affordance-to-action
ICLR 2022VAT-MartVAT-Mart: Learning Visual Action Trajectory Proposals for Manipulating 3D Articulated ObjectsPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2023HULC++Grounding Language with Visual Affordances over Unstructured DataPaperE2EimplicitIntegrated policy · Implicit representation
ICLR 2023DualAffordDualAfford: Learning Collaborative Visual Affordance for Dual-gripper Object ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2023RLAffordRLAfford: End-to-End Affordance Learning for Robotic ManipulationPaperE2EimplicitIntegrated policy · Implicit representation
ICRA 2023-Visual Affordance Prediction for Guiding Robot ExplorationPaperE2ErewardIntegrated policy · Optimization signal
CoRL 2023ACE-NBVAffordance-Driven Next-Best-View Planning for Robotic GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
NeurIPS 2023-Learning Environment-aware Affordance for 3D Articulated Object Manipulation under OcclusionsPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2023LERF-TOGOLanguage Embedded Radiance Fields for Zero-Shot Task-Oriented GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2023VoxPoserVoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language ModelsPaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2023VRBAffordances from Human Videos as a Versatile Representation for RoboticsPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2023DefoAffordLearning Foresightful Dense Visual Affordance for Deformable Object ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IEEE T-RO 2024PRIMPPRIMP: Probabilistically-Informed Motion Primitives for Efficient Affordance Learning from DemonstrationPaperMotion primitivesexplicitHierarchical affordance-to-action · Primitive-based
arXiv 2024AnyPartOpen-Vocabulary Part-Based GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2024IDAInformation-driven Affordance Discovery for Efficient Robotic ManipulationPaperHierarchical modelrewardHierarchical affordance-to-action · Planner-based
arXiv 2024-Affordance-based Robot Manipulation with Flow MatchingPaperE2EexplicitIntegrated policy · Explicit input
CoRL 2024ReKepReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ECCV 2024Robo-ABCRobo-ABC: Affordance Generalization Beyond Categories via Semantic Correspondence for Robot ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IROS 2024PreAffordPreAfford: Universal Affordance-Based Pre-Grasping for Diverse Objects and EnvironmentsPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2024RAMRAM: Retrieval-Based Affordance Transfer for Generalizable Zero-Shot Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2025GATLearning Precise Affordances from Egocentric Videos for Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICRA 2025RT-AffordanceRT-Affordance: Affordances are Versatile Intermediate Representations for Robot ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IROS 2025ManipGPTManipGPT: Is Affordance Segmentation by Large Vision Models Enough for Articulated Object Manipulation?PaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2025AffordDexGraspAffordDexGrasp: Open-set Language-guided Dexterous Grasp with Generalizable-Instructive AffordancePaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2025GarmentPileGarmentPile: Point-Level Visual Affordance Guided Retrieval and Adaptation for Cluttered Garments ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
ICCV 2025A0A0: An Affordance-Aware Hierarchical Model for General Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IEEE CBS 2025DORADORA: Object Affordance-Guided Reinforcement Learning for Dexterous Robotic ManipulationPaperE2ErewardIntegrated policy · Optimization signal
CoRL 2025GLOVER++GLOVER++: Unleashing the Potential of Affordance Learning from Human Behaviors for Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2025AffordDPAffordDP: Generalizable Diffusion Policy with Transferable AffordancePaperE2EexplicitIntegrated policy · Explicit input
ICML 2025BiAssembleBiAssemble: Learning Collaborative Affordance for Bimanual Geometric AssemblyPaperHierarchical modelexplicitHierarchical affordance-to-action
CoRL 2025O3AffordO3Afford: One-Shot 3D Object-to-Object Affordance Grounding for Generalizable Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IEEE/ASME Transactions on Mechatronics 2025DARTDART: A Novel Task-Driven Diffusion-Based Policy with Affordance Learning for Generalizable Manipulation of Articulated ObjectsPaperE2EexplicitIntegrated policy · Explicit input
ICCV 2025CoA-VLACoA-VLA: Improving Vision-Language-Action Models via Visual-Textual Chain-of-AffordancePaperE2EexplicitIntegrated policy · Explicit input
ICCV 20252HandedAfforder2HandedAfforder: Learning Precise Actionable Bimanual Affordances from Human VideosPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2025OVAL-GraspOVAL-Grasp: Open-Vocabulary Affordance Localization for Task Oriented GraspingPaperHierarchical modelexplicitHierarchical affordance-to-action
RA-L 2025ScaleADFGScaleADFG: Affordance-based Dexterous Functional Grasping via Scalable DatasetPaperHierarchical modelexplicitHierarchical affordance-to-action
AAAI 2026AffordDexTowards Affordance-Aware Robotic Dexterous Grasping with Human-like PriorsPaperE2ErewardIntegrated policy · Optimization signal
CVPR 2026AFIAffordance Field Intervention: Enabling VLAs to Escape Memory Traps in Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
AAAI 2026A3DA3D: Adaptive Affordance Assembly with Dual-Arm ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
IROS 2026FSAGFSAG: Enhancing Human-to-Dexterous-Hand Finger-Specific Affordance Grounding via Diffusion ModelsPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026CorDexGenerate, Transfer, Adapt: Learning Functional Dexterous Grasping from a Single Human DemonstrationPaperHierarchical modelimplicitHierarchical affordance-to-action · Retrieval-based
IEEE Access 2026AffoRo-GSOne-Shot 3-D Affordance Learning for Multi-Stage Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action
CVPR 2026BiPreManipBiPreManip: Learning Affordance-Based Bimanual Preparatory Manipulation through Anticipatory CollaborationPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026RAAPRAAP: Retrieval-Augmented Affordance Prediction with Cross-Image Action AlignmentPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026BridgeACTBridgeACT: Bridging Human Demonstrations to Robot Actions via Unified Tool-Target AffordancesPaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026Afford-VLAAfford-VLA: Action-Aligned Visual Planning via Internalized AffordancePaperE2EimplicitIntegrated policy · Implicit representation
arXiv 2026AffordVLAAffordVLA: Injecting Affordance Representations into Vision-Language-Action Models via Implicit Feature AlignmentPaperE2EimplicitIntegrated policy · Implicit representation
CVPR 2026PALMPALM: Progress-Aware Policy Learning via Affordance Reasoning for Long-Horizon Robotic ManipulationPaperE2EimplicitIntegrated policy · Implicit representation
CVPR 2026AffordGenAffordGen: Generating Diverse Demonstrations for Generalizable Object Manipulation with Affordance CorrespondencePaperE2EimplicitIntegrated policy · Implicit representation
arXiv 2026AffordanceVLAAffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware UnderstandingPaperE2EimplicitIntegrated policy · Implicit representation
arXiv 2026Affordance2ActionAffordance2Action: Task-Conditioned Scene-level Affordance Grounding for Real-Time ManipulationPaperE2Eexplicit/implicitIntegrated policy · Explicit/implicit representation
arXiv 2026GROW2GROW2: Grounding Which and Where for Robot Tool UsePaperHierarchical modelexplicitHierarchical affordance-to-action
arXiv 2026AffordSimAffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic ManipulationPaperHierarchical modelexplicitHierarchical affordance-to-action · Simulation benchmark

📊 Affordance Datasets / Benchmarks

Survey groups: object-centric · scene-level · interaction-driven · language- and reasoning-oriented · action-oriented.

Venue/DateNameTitlePaperData Modality
ICRA 2015UMD RGB-D Part Affordance DatasetAffordance Detection of Tool Parts from Geometric FeaturesPaperimage
ECCV 2016NYUv2 affordance annotationsA Multi-scale CNN for Affordance Segmentation in RGB ImagesPaperimage
IROS 2017IIT-AFFObject-Based Affordances Detection with Convolutional Neural Networks and Dense Conditional Random FieldsPaperimage
CVPR 2018ADE-AffordanceLearning to Act Properly: Predicting and Explaining Affordances From ImagesPaperimage
CVPR 2018OPRADemo2Vec: Reasoning Object Affordances from Online VideosPapervideo-image
ECCV 2018EPIC-KITCHENSScaling egocentric vision: The epic-kitchens datasetPapervideo
CVPR 2020GraspNet-1BillionGraspnet-1billion: A large-scale benchmark for general object graspingPaperRGB-D/3D grasp
CVPR 20213D AffordanceNet3D AffordanceNet: A Benchmark for Visual Object Affordance UnderstandingPaper3D
CVPR 2022AGD20KLearning Affordance Grounding from Exocentric ImagesPaperimage-image
CVPR 2022Ego4DEgo4d: Around the world in 3,000 hours of egocentric videoPapervideo
ECCVW 2022PartAffordPartAfford: Part-level Affordance Discovery from 3D ObjectsPaper3D
ICCV 20233DOIUnderstanding 3D Object Interaction from a Single ImagePaperimage/3D interaction
ICCV 2023AffordPoseAffordPose: A Large-Scale Dataset of Hand-Object Interactions with Affordance-Driven Hand PosePaper3D
ICCV 2023EPIC-AffMulti-label Affordance Mapping from Egocentric VisionPapervideo-3D
ICCV 2023PIADGrounding 3D Object Affordance from 2D Interactions in ImagesPaperimage-3D
IROS 2023HANDALHANDAL: A Dataset of Real-World Manipulable Object Categories with Pose Annotations, Affordances, and ReconstructionsPaperRGB-D/6-DoF pose/3D mesh
WACV 2023EPIC-KITCHENS AffordanceFine-Grained Affordance Annotation for Egocentric Hand-Object Interaction VideosPapervideo/image
CVPR 20243DIRLEMON: Learning 3D Human-Object Interaction Relation from 2D ImagesPaperimage-3D
CVPR 2024AffordQLASO: Language-guided Affordance Segmentation on 3D ObjectPaper3D
CVPR 2024SceneFun3DSceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D ScenesPaper3D
ACL 2025InstructPartInstructPart: Task-Oriented Part Segmentation with Instruction ReasoningPaperimage-language
ACM MM 20253DAffordSplat3DAffordSplat: Efficient Affordance Reasoning with 3D GaussiansPaper3D
ACM MM 2025RoboAffordRoboAfford: A Dataset and Benchmark for Enhancing Object and Spatial Affordance Learning in Robot ManipulationPaperimage/language
arXiv 2025A4BenchAffordance Benchmark for MLLMsPaperimage
arXiv 2025Affogato-750KAffogato: Learning Open-Vocabulary Affordance Grounding with Automated Data Generation at ScalePaperlanguage-3D
arXiv 2025HOVA-500KGLOVER++: Unleashing the Potential of Affordance Learning from Human Behaviors for Robotic ManipulationPaperimage
arXiv 2025SeqAffordSplatSeqAffordSplat: Scene-level Sequential Affordance Reasoning on 3D Gaussian SplattingPaperlanguage/3D Gaussian
CVPR 2025AGPILGrounding 3D Object Affordance with Language Instructions, Visual Observations and InteractionsPaperimage-3D
CVPR 2025FunGraph3DOpen-Vocabulary Functional 3D Scene Graphs for Real-World Indoor SpacesPaperRGB-D/3D scene graph
CVPR 2025PIADv2GREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance GroundingPaperimage-language-3D
CVPR 2025SeqAffordSeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language ModelPaperlanguage-3D
ICCV 20252HandedAffordance2HandedAfforder: Learning Precise Actionable Bimanual Affordances from Human VideosPapervideo/text/3D
ICCV 2025RAGNetRAGNet: Large-scale Reasoning-based Affordance Segmentation Benchmark towards General GraspingPaperimage
IROS 2025AGDDO15One-Shot Affordance Grounding of Deformable Objects in Egocentric Organizing ScenesPaperimage
IROS Workshop 2025RoboAfford++RoboAfford++: A Generative AI-Enhanced Dataset for Multimodal Affordance Learning in Robotic Manipulation and NavigationPaperimage/language
RA-L 2025ScaleADFGScaleADFG: Affordance-based Dexterous Functional Grasping via Scalable DatasetPaper3D/grasp
AAAI 2026ReasonAffAffordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language ModelPaperimage/language
arXiv 2026360-AGDPanoAffordanceNet: Towards Holistic Affordance Grounding in 360° Indoor EnvironmentsPaperpanoramic image/text
arXiv 2026A2A-BenchAffordance2Action: Task-Conditioned Scene-level Affordance Grounding for Real-Time ManipulationPaperimage
arXiv 2026AffordSimAffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic ManipulationPapersimulation/3D/robot action
arXiv 2026CompassADCompassAD: Intent-Driven 3D Affordance Grounding in Functionally Competing ObjectsPaperlanguage-3D
arXiv 2026VIDAVideoAfford: Grounding 3D Affordance from Human-Object-Interaction Videos via Multimodal Large Language ModelPapervideo-3D
CVPR 2026AffordBridgeAffordMatcher: Affordance Learning in 3D Scenes from Visual SignifiersPaperimage-3D
ECCV 2026PAP-12KPanoramic Affordance PredictionPaperpanoramic image/text

📚 Related Surveys

Venue/DateNameTitlePaperFocus
IEEE TCDS 2016-Affordance Research in Developmental Robotics: A SurveyPaperDevelopmental robotics and robot acquisition of affordance knowledge
IEEE TCDS 2018-Affordances in Psychology, Neuroscience, and Robotics: A SurveyPaperCross-disciplinary foundations spanning psychology, neuroscience, and robotics
IJCAI 2021-Building Affordance Relations for Robotic Agents - A ReviewPaperRelational affordance representations for robotic agents
ACM Computing Surveys 2021-Visual Affordance and Function Understanding: A SurveyPaperVisual affordance recognition, segmentation, and function understanding
IEEE TBD 2023-A Survey of Visual Affordance Recognition Based on Deep LearningPaperDeep-learning methods for visual affordance recognition
IEEE TCDS 2023-Recent Advances of Deep Robotic Affordance Learning: A Reinforcement Learning PerspectivePaperRobotic affordance learning from a reinforcement-learning perspective
arXiv 2025-Visual Affordance Prediction: Survey and ReproducibilityPaperVisual affordance prediction with an emphasis on reproducibility

🎉 Contributing

⭐ Help us grow this repository! If you know any valuable works we’ve missed, don’t hesitate to contribute — every suggestion makes a difference!

We welcome and appreciate all contributions! Here’s how you can help:

  • 📄 Add or Update a Paper
    Contribute by adding a new paper or improving details of an existing one. Please consider the most appropriate category for the work.

  • ✍️ Use Consistent Formatting
    Follow the format of the existing entries to maintain clarity and consistency across the list.

  • 🔗 Include Abstract Link
    If the paper is from arXiv, use the /abs/ link format for the abstract (e.g., https://arxiv.org/abs/xxxx.xxxxx).

  • 💡 Explain Your Edit (Optional but Helpful)
    A short note on why you think the paper deserves to be added or updated is appreciated and helps maintainers process your PR faster.

✅ Don't worry about getting everything perfect!
Minor mistakes are totally fine — we’ll help fix them. What matters most is your contribution. Let's highlight your awesome work together!

🌟 Acknowledgment

Thanks for the wonderful researchers focusing on affordance learning and embodied AI

📄 License

This project is licensed under the MIT License.

Similar repos

YanjieZe/Paper-List

knowledge

603

BaiShuanghao/Awesome-Robotics-Manipulation

A comprehensive list of papers about Robot Manipulation, including papers, codes, and related…

1,151

lhc1224/Cross-View-AG

Official PyTorch Implementation of Learning Affordance Grounding from Exocentric Images, CVPR 2022

77

robotics-survey/Awesome-Robotics-Foundation-Models

1,411

Sid2697/awesome-egocentric-vision

A curated list of egocentric (first-person) vision and related area resources

348

yueen-ma/Awesome-VLA

671

DoranLyong/oh-my-world-intelligence

From 3D/4D Scene Understanding to Semantic-metric World Models

0

showlab/Awesome-Multimodal-Embodied-Agent

Survey on Multimodal Embodied Agents: From Computer-Use to Robot-Use

70

Contributors

hq-King

64 commits

Zhenhao-Zhang

2 commits

ha8siouhdi

1 commits

XubeiPan666

1 commits