This is the official repository of the paper "Object size prediction from hand movement using a single RGB sensor". It contains the implementation, the dataset and the supplemental material of our work.

CHAIR: 3D Human-Interaction Dataset. Human Object Interaction (HOI) dataset, focusing on interactions between humans and chairs.

Mask2IV: Interaction-Centric Video Generation via Mask Trajectories (AAAI 2026)

3D Human-Object Interaction in Video A New Approach to Object Tracking via Cross-Modal Attention

One-Stage HOI detection code: Visual Compositional Learning for Human-Object Interaction Detection (ECCV2020), Affordance Transfer Learning for Human-Object Interaction Detection (CVPR2021)

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

A toolkit developed to help analyzing existing Human-Object Interaction (HOI) detection methods, for research purpose. Not maintained but might still work for the majority of HOI methods..

A toolbox for HOI Detection.

Utilities for the dataset V-COCO in Python3

This is the repo for an open project named detecting human object interactions in real-time

[AAAI 2022] Detecting Human-Object Interactions with Object-Guided Cross-Modal Calibrated Semantics.

[ICCV'25] Method for generating static human-object interactions

This repository contains the dataset and the source code for the detection of visual relationships with the Logic Tensor Networks framework.

Code for the ECCV'22 paper "Geometric Features Informed Multi-person Human-object Interaction Recognition in Videos".

🚴‍♂️ ConsNet: Learning Consistency Graph for Zero-Shot Human-Object Interaction Detection (MM 2020)

Code for our paper "Category Query Learning for Human-Object Interaction Classification" (CVPR2023)

LOCATE: Localize and Transfer Object Parts for Weakly Supervised Affordance Grounding (CVPR 2023)

[CVPR'23] Method to infer objects from human interactions

Code for CVPR22 paper: Exploring Structure-aware Transformer over Interaction Proposals for Human-Object Interaction Detection.

The official code for Relational Context Learning for Human-Object Interaction Detection, CVPR2023.