NARF22: Neural Articulated Radiance Fields for Configuration-Aware Rendering
This addresses scalability issues in robotic manipulation by enabling efficient rendering of articulated objects, though it is incremental as it builds on existing NeRF methods.
The paper tackles the challenge of rendering articulated objects for robotic perception by proposing Neural Articulated Radiance Fields (NARF22), a pipeline that uses a differentiable, configuration-parameterized NeRF to provide high-quality renderings, requiring no explicit object structure knowledge at inference and generalizing well with minimal training data.
Articulated objects pose a unique challenge for robotic perception and manipulation. Their increased number of degrees-of-freedom makes tasks such as localization computationally difficult, while also making the process of real-world dataset collection unscalable. With the aim of addressing these scalability issues, we propose Neural Articulated Radiance Fields (NARF22), a pipeline which uses a fully-differentiable, configuration-parameterized Neural Radiance Field (NeRF) as a means of providing high quality renderings of articulated objects. NARF22 requires no explicit knowledge of the object structure at inference time. We propose a two-stage parts-based training mechanism which allows the object rendering models to generalize well across the configuration space even if the underlying training data has as few as one configuration represented. We demonstrate the efficacy of NARF22 by training configurable renderers on a real-world articulated tool dataset collected via a Fetch mobile manipulation robot. We show the applicability of the model to gradient-based inference methods through a configuration estimation and 6 degree-of-freedom pose refinement task. The project webpage is available at: https://progress.eecs.umich.edu/projects/narf/.