iMODE:Real-Time Incremental Monocular Dense Mapping Using Neural Field
Hidenobu Matsuki, Edgar Sucar, Tristan Laidow, Kentaro Wada, Raluca Scona, Andrew J. Davison
Abstract
We present a novel real-time dense and semantic neural field mapping system that uses only monocular images as input. Our scene representation is a dense continuous radiance field represented by a Multi-Layer Perceptron (MLP), trained from scratch in real-time. We build on high-performance sparse visual SLAM and use camera poses and sparse keypoint depths as supervision alongside RGB keyframes. Since no prior training is required, our system flexibly fits to arbitrary scale and structure at runtime, and works even with strong specular reflections. We demonstrate reconstruction over a range of scenes from small indoor to large outdoor spaces. We also show that the method can straightforwardly benefit from additional inputs such as learned depth priors or semantic labels for more precise and advanced mapping.
BibTeX
@inproceedings{icra2023_imoderealtimeinc,
title = {iMODE:Real-Time Incremental Monocular Dense Mapping Using Neural Field},
author = {Hidenobu Matsuki and Edgar Sucar and Tristan Laidow and Kentaro Wada and Raluca Scona and Andrew J. Davison},
booktitle = {ICRA 2023},
year = {2023}
}