Multi-Source Domain Fusion Cross-Domain Pedestrian Recognition Based on High-Quality Intermediate Domains
Lists
Only you see this
Abstract
Pedestrian detection has received considerable attention over the last few years because the technology can be combined with pedestrian tracking and re-identification in areas such as vehicle-assisted driving and intelligent video surveillance. However, existing pedestrian detection techniques have achieved excellent results. Problems such as domain gaps lead to poor generalization performance, thus limiting its application and practical value. This paper proposes a High-Quality Integration Domain framework for pedestrian recognition. First, the source domains are produced as super-resolution training data. The HCycleGAN model uses super-resolution algorithms and a generative framework to generate high-quality intermediate domains. Second, a multi-source domain fusion scheme based on the NPIQE module is proposed to improve the generated framework’s quality and reduce the dataset’s overfitting. It fuses images by three aspects: similarity, blurriness and unsupervised image quality score values. Finally, we use an anchor-free Center and Scale Prediction model for pedestrian detection. The experimental dataset contains two common pedestrian detection datasets, Caltech and CityPersons. Cross-domain experimental results show that the framework can improve the cross-domain detection from CityPersons to Caltech by 6% and from Caltech to CityPersons by 5%. The training of CityPersons to Caltech achieves almost the same detection accuracy as that of the Caltech original domain. In conclusion, the framework in this paper is effective for cross-domain pedestrian detection and can provide ideas and inspiration for future practical applications.
BibTeX
@inproceedings{niu2023multi,
author = {Yixing Niu and Wansheng Cheng and Y. F. Lai and Hongzhi Zhang and Mingrui Cao and Kang Cao and Song Fan},
title = {{Multi-Source Domain Fusion Cross-Domain Pedestrian Recognition Based on High-Quality Intermediate Domains}},
booktitle = {IEEE Access},
year = {2023},
doi = {10.1109/access.2023.3297265},
}Loading related content...