STAIR: Semantic-Targeted Active Implicit Reconstruction
This work addresses the need for object-level understanding in robotics, enabling more efficient reconstruction of semantically relevant objects in unknown environments, though it appears incremental as it builds on existing implicit neural representation methods.
The authors tackled the problem of actively reconstructing objects with specific semantic meanings for autonomous robots by proposing a framework using semantic implicit neural representations and adaptive view planning. Their approach achieved better reconstruction performance in mesh and novel view rendering quality compared to baselines and outperformed a state-of-the-art explicit map-based pipeline.
Many autonomous robotic applications require object-level understanding when deployed. Actively reconstructing objects of interest, i.e. objects with specific semantic meanings, is therefore relevant for a robot to perform downstream tasks in an initially unknown environment. In this work, we propose a novel framework for semantic-targeted active reconstruction using posed RGB-D measurements and 2D semantic labels as input. The key components of our framework are a semantic implicit neural representation and a compatible planning utility function based on semantic rendering and uncertainty estimation, enabling adaptive view planning to target objects of interest. Our planning approach achieves better reconstruction performance in terms of mesh and novel view rendering quality compared to implicit reconstruction baselines that do not consider semantics for view planning. Our framework further outperforms a state-of-the-art semantic-targeted active reconstruction pipeline based on explicit maps, justifying our choice of utilising implicit neural representations to tackle semantic-targeted active reconstruction problems.