Engelhardt H, Kalenberg M, Franke J, Martin S (2025)
Publication Type: Conference contribution
Publication year: 2025
Publisher: Springer Science and Business Media Deutschland GmbH
Book Volume: 2629 CCIS
Pages Range: 348-363
Conference Proceedings Title: Communications in Computer and Information Science
Event location: Porto, PRT
ISBN: 9783032009852
DOI: 10.1007/978-3-032-00986-9_23
This paper presents Learn Where I Can Walk (LWICW), a novel auto-labeling approach for segmenting walked areas using a trajectory estimated from multiple sequential monocular camera images, aimed at training supervised segmentation models for the navigation of visually impaired people. The proposed method uses images sourced from Mapillary, which is a collaborative platform to share street-level images. The approach involves extracting the walked path of the camera operator through the camera poses, the filtering of occluded walking path poses using Depth Anything V2, and the application of Segment Anything Model 2 (SAM 2) for segmentation. The LWICW auto-labels are validated against a manually labeled dataset from Mapillary and compared to the state-of-the-art zero-shot segmentation model Grounded SAM 2. The LWICW method achieves an overall mean Intersection over Union (mIoU) of 93.9% and a mean F1 score (mF1) of 96.6%, which represents a performance improvement of +1.6% points on mIoU and +1.0% points on mF1 compared to the Grounded SAM 2 approach.
APA:
Engelhardt, H., Kalenberg, M., Franke, J., & Martin, S. (2025). Learn Where I Can Walk: Auto-labeling of Walked Areas Using Monocular Camera Trajectory. In Juha Röning, Joaquim Filipe (Eds.), Communications in Computer and Information Science (pp. 348-363). Porto, PRT: Springer Science and Business Media Deutschland GmbH.
MLA:
Engelhardt, Helmut, et al. "Learn Where I Can Walk: Auto-labeling of Walked Areas Using Monocular Camera Trajectory." Proceedings of the 5th International Conference on Robotics, Computer Vision and Intelligent Systems, ROBOVIS 2025, Porto, PRT Ed. Juha Röning, Joaquim Filipe, Springer Science and Business Media Deutschland GmbH, 2025. 348-363.
BibTeX: Download