vMAP: Vectorised Object Mapping for Neural Field SLAM
This work addresses object-level mapping in SLAM for robotics or AR/VR applications, offering incremental improvements in efficiency and quality over existing neural field methods.
The authors tackled the problem of dense SLAM using neural fields by introducing vMAP, which represents each object with a small MLP for efficient, watertight modeling without 3D priors, achieving a training speed of 5Hz map updates and improved reconstruction quality compared to prior systems.
We present vMAP, an object-level dense SLAM system using neural field representations. Each object is represented by a small MLP, enabling efficient, watertight object modelling without the need for 3D priors. As an RGB-D camera browses a scene with no prior information, vMAP detects object instances on-the-fly, and dynamically adds them to its map. Specifically, thanks to the power of vectorised training, vMAP can optimise as many as 50 individual objects in a single scene, with an extremely efficient training speed of 5Hz map update. We experimentally demonstrate significantly improved scene-level and object-level reconstruction quality compared to prior neural field SLAM systems. Project page: https://kxhit.github.io/vMAP.