Structure from Action: Learning Interactions for Articulated Object 3D Structure Discovery
This addresses the challenge of constructing 3D articulated CAD models for object categories not seen during training, which is incremental as it builds on existing interaction and perception methods.
The paper tackles the problem of discovering 3D part geometry and joint parameters for unseen articulated objects by using inferred interactions, achieving a 25.4 percentage point improvement in 3D IoU over state-of-the-art components on unseen categories.
We introduce Structure from Action (SfA), a framework to discover 3D part geometry and joint parameters of unseen articulated objects via a sequence of inferred interactions. Our key insight is that 3D interaction and perception should be considered in conjunction to construct 3D articulated CAD models, especially for categories not seen during training. By selecting informative interactions, SfA discovers parts and reveals occluded surfaces, like the inside of a closed drawer. By aggregating visual observations in 3D, SfA accurately segments multiple parts, reconstructs part geometry, and infers all joint parameters in a canonical coordinate frame. Our experiments demonstrate that a SfA model trained in simulation can generalize to many unseen object categories with diverse structures and to real-world objects. Empirically, SfA outperforms a pipeline of state-of-the-art components by 25.4 3D IoU percentage points on unseen categories, while matching already performant joint estimation baselines.