InteractMove is a dataset for generating human-object interactions in 3D scenes, released in 2025 by the Institute of Computer Science and Technology at Peking University in collaboration with the Beijing Institute of Electronic Science and Technology. The related research paper is titled "InteractMove: Text-Controlled Human-Object Interaction Generation in 3D Scenes with Movable ObjectsThe aim is to support and promote research on text-based interactive modeling of movable objects.

This dataset covers multiple types of movable objects and various real-world scanning scenes, and provides human-object interaction sequences that are strictly aligned with the scenes. The overall data structure can be divided into the following three parts:

• 3D Scenes: Real-world indoor scenes from ScanNet V2, ranging from scene0000_00 to scene0706_00, providing complex environments with multiple objects, multiple occlusions, and multiple rooms.
• Aligned Motions: These are sequences of human-object interactions aligned with the scene, including human pose, hand information, and object trajectory.
• Movable Object Meshes: The object models come from the GRAB and BEHAVE datasets, covering multiple types of interactive objects. After unified processing, they are embedded into various scenes for action interaction.

This project only provides the second part (Aligned Motions); the remaining data is not included.