A Cross-Season Correspondence Dataset for Robust Semantic Segmentation
Paper in proceedings, 2019

In this paper, we present a method to utilize 2D-2D point matches between images taken during different image conditions to train a convolutional neural network for semantic segmentation. Enforcing label consistency across the matches makes the final segmentation algorithm robust to seasonal changes. We describe how these 2D-2D matches can be generated with little human interaction by geometrically matching points from 3D models built from images. Two cross-season correspondence datasets are created providing 2D-2D matches across seasonal changes as well as from day to night. The datasets are made publicly available to facilitate further research. We show that adding the correspondences as extra supervision during training improves the segmentation performance of the convolutional neural network, making it more robust to seasonal changes and weather conditions.

Author

Måns Larsson

Chalmers, Electrical Engineering, Signalbehandling och medicinsk teknik, Imaging and Image Analysis

Erik Stenborg

Chalmers, Electrical Engineering, Signalbehandling och medicinsk teknik, Signal Processing

Lars Hammarstrand

Chalmers, Electrical Engineering, Signalbehandling och medicinsk teknik, Signal Processing

Marc Pollefeys

Torsten Sattler

Chalmers, Electrical Engineering, Signalbehandling och medicinsk teknik, Imaging and Image Analysis

Fredrik Kahl

Chalmers, Electrical Engineering, Signalbehandling och medicinsk teknik, Imaging and Image Analysis

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

10636919 (ISSN)

IEEE International Conference on Computer Vision (CVPR) 2019
Long Beach, USA,

COPPLAR CampusShuttle cooperative perception & planning platform

VINNOVA, 2016-01-01 -- 2018-12-31.

Subject Categories

Computer Vision and Robotics (Autonomous Systems)

More information

Created

5/14/2019