ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume V-2-2021
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-2-2021, 83–91, 2021
https://doi.org/10.5194/isprs-annals-V-2-2021-83-2021
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-2-2021, 83–91, 2021
https://doi.org/10.5194/isprs-annals-V-2-2021-83-2021

  17 Jun 2021

17 Jun 2021

SEMI-SUPERVISED SEGMENTATION OF CONCRETE AGGREGATE USING CONSENSUS REGULARISATION AND PRIOR GUIDANCE

M. Coenen1, T. Schack1, D. Beyer1, C. Heipke2, and M. Haist1 M. Coenen et al.
  • 1Institute of Building Materials Science, Leibniz University Hannover, Germany
  • 2Institute of Photogrammetry and GeoInformation, Leibniz University Hannover, Germany

Keywords: semi-supervised learning, semantic segmentation, consistency training, auto-encoder, concrete aggregate particles

Abstract. In order to leverage and profit from unlabelled data, semi-supervised frameworks for semantic segmentation based on consistency training have been proven to be powerful tools to significantly improve the performance of purely supervised segmentation learning. However, the consensus principle behind consistency training has at least one drawback, which we identify in this paper: imbalanced label distributions within the data. To overcome the limitations of standard consistency training, we propose a novel semi-supervised framework for semantic segmentation, introducing additional losses based on prior knowledge. Specifically, we propose a lightweight architecture consisting of a shared encoder and a main decoder, which is trained in a supervised manner. An auxiliary decoder is added as additional branch in order to make use of unlabelled data based on consensus training, and we add additional constraints derived from prior information on the class distribution and on auto-encoder regularisation. Experiments performed on our concrete aggregate dataset presented in this paper demonstrate the effectiveness of the proposed approach, outperforming the segmentation results achieved by purely supervised segmentation and standard consistency training.