Aff-Grasp

KO | EN

gitlite — search

#claude-code #ai #developer-tools #claude #ai-agent #cli #mcp #macos #python #ai-agents #codex #llm

★ 26 Open GitHub ↗

Learning Precise Affordances from Egocentric Videos for Robotic Manipulation (ICCV 2025)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Learning Precise Affordances from Egocentric Videos for Robotic Manipulation

[![arXiv](https://img.shields.io/badge/arXiv-2408.10123-b31b1b.svg)](https://arxiv.org/abs/2408.10123)
[![GitHub](https://img.shields.io/website?label=Project%20&up_message=website&url=https://reagan1311.github.io/affgrasp/)](https://reagan1311.github.io/affgrasp/)

## Overview
- Code for affordance extraction from egocentric videos is in the folder `ego2aff`
- Code for affordance model learning is in the folder `affordance-learning`
- Data: [Data_for_Aff-Grasp](https://huggingface.co/datasets/Gen1113/Data_for_Aff-Grasp)
- Model and Log: [Model_for_Aff-Grasp](https://huggingface.co/Gen1113/Model_for_Aff-Grasp)

## Citation

```
@article{li2025affgrasp,
      title     = {Learning Precise Affordances from Egocentric Videos for Robotic Manipulation}, 
      author    = {Li, Gen and Tsagkas, Nikolaos and Song, Jifei and Mon-Williams, Ruaridh and Vijayakumar, Sethu and Shao, Kun and Sevilla-Lara, Laura},
      journal   = {Proceedings of the IEEE/CVF International Conference on Computer Vision},
      year      = {2025},
    }
```

## Anckowledgement
Part of the code is derived from the [hand_object_detector](https://github.com/ddshan/hand_object_detector), [hoi-forecase](https://github.com/stevenlsw/hoi-forecast), [GroundedSAM](https://github.com/IDEA-Research/Grounded-Segment-Anything), and [ViT-Adapter](https://github.com/czczup/ViT-Adapter). Thanks for their great work!

Aff-Grasp

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README