FreeLoc enables accurate indoor localization using floorplans and RGB images
FreeLoc: Online Floorplan Localization via Diffusion-Aided Pose Refinement
RoboticsComputer Vision and Pattern Recognition
Summary
Indoor robots and devices often need to figure out their exact location inside a building. Floorplans are simple maps of buildings, but existing methods require heavy setup and storage to match real-world positions. The authors present FreeLoc, a system that uses camera images and floorplans directly, without needing large pre-made databases. It quickly guesses and refines positions in real time, making indoor localization faster and more practical for robots.
What this means in practice
- •For indoor robotics teams: Implement real-time robot self-localization inside buildings using existing floorplans and onboard RGB cameras without needing pre-built databases.
- •For facility management teams: Enhance indoor asset tracking and navigation systems by integrating on-the-fly localization based on floorplans and camera inputs.
Authors
Haocheng Peng, Boyang Zhou, Jiarui Hu, Xiyue Guo, Ziyang Zhang, Boming Zhao, Yifan Gao, Xiao Li, Hujun Bao, Zhaopeng Cui
Abstract
Floorplans provide compact and widely available geometric maps for indoor localization, but existing high-performing floorplan-based methods still convert them into dense scene-specific offline databases, tying accuracy, storage, and runtime to the sampling resolution of the discretized pose space. We present FreeLoc, an online RGB-based floorplan localization framework that treats the floorplan as a directly queryable geometric map. FreeLoc introduces an efficient online geometric querying and diffusion-aided refinement scheme, which retrieves plausible pose anchors through on-the-fly floorplan ray querying and refines them into accurate continuous pose estimates. For sequential localization, FreeLoc develops an online likelihood construction strategy that bridges single-frame localization and probabilistic temporal fusion by constructing likelihoods from coarse-sampled candidates and refined pose hypotheses, enabling histogram-filter-based temporal fusion without offline databases. Experiments demonstrate real-time online inference and state-of-the-art performance in both single-frame and sequential localization, while real-world results validate practical deployability in indoor robotic localization scenarios.