Header

UZH-Logo

Maintenance Infos

Self-supervised Human Detection and Segmentation via Background Inpainting


Katircioglu, Isinsu; Rhodin, Helge; Constantin, Victor; Spörri, Jörg; Salzmann, Mathieu; Fua, Pascal (2022). Self-supervised Human Detection and Segmentation via Background Inpainting. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(12):9574-9588.

Abstract

While supervised object detection and segmentation methods achieve impressive accuracy, they generalize poorly to images whose appearance significantly differs from the data they have been trained on. To address this when annotating data is prohibitively expensive, we introduce a self-supervised detection and segmentation approach that can work with single images captured by a potentially moving camera. At the heart of our approach lies the observation that object segmentation and background reconstruction are linked tasks, and that, for structured scenes, background regions can be re-synthesized from their surroundings, whereas regions depicting the moving object cannot. We encode this intuition into a self-supervised loss function that we exploit to train a proposal-based segmentation network. To account for the discrete nature of the proposals, we developed a Monte Carlo-based training strategy that allows the algorithm to explore the large space of object proposals. We apply our method to human detection and segmentation in images that visually depart from those of standard benchmarks and outperform existing self-supervised methods.

Abstract

While supervised object detection and segmentation methods achieve impressive accuracy, they generalize poorly to images whose appearance significantly differs from the data they have been trained on. To address this when annotating data is prohibitively expensive, we introduce a self-supervised detection and segmentation approach that can work with single images captured by a potentially moving camera. At the heart of our approach lies the observation that object segmentation and background reconstruction are linked tasks, and that, for structured scenes, background regions can be re-synthesized from their surroundings, whereas regions depicting the moving object cannot. We encode this intuition into a self-supervised loss function that we exploit to train a proposal-based segmentation network. To account for the discrete nature of the proposals, we developed a Monte Carlo-based training strategy that allows the algorithm to explore the large space of object proposals. We apply our method to human detection and segmentation in images that visually depart from those of standard benchmarks and outperform existing self-supervised methods.

Statistics

Citations

Dimensions.ai Metrics
3 citations in Web of Science®
3 citations in Scopus®
Google Scholar™

Altmetrics

Additional indexing

Item Type:Journal Article, refereed, original work
Communities & Collections:04 Faculty of Medicine > Balgrist University Hospital, Swiss Spinal Cord Injury Center
Dewey Decimal Classification:610 Medicine & health
Scopus Subject Areas:Physical Sciences > Software
Physical Sciences > Computer Vision and Pattern Recognition
Physical Sciences > Computational Theory and Mathematics
Physical Sciences > Artificial Intelligence
Physical Sciences > Applied Mathematics
Language:English
Date:1 December 2022
Deposited On:17 Jan 2022 09:21
Last Modified:26 Apr 2024 01:39
Publisher:Institute of Electrical and Electronics Engineers
ISSN:0098-5589
OA Status:Closed
Publisher DOI:https://doi.org/10.1109/TPAMI.2021.3123902
PubMed ID:34714741
Full text not available from this repository.