Communication-efficient Multi-source Domain Adaptive Object Detection under Privacy Constraints

Peggy Joy Lu*, Jen Hui Chuang

*此作品的通信作者

研究成果: Conference contribution同行評審

摘要

To establish a more generalized model in object detection, collaboration among multiple cameras can increase data diversity and reduce the effort for data collection, leading to a new research area as multi-source domain adaptative object detection (MSDAOD). However, preserving source data privacy in MSDAOD is challenging due to the lack of information integrated from all source domains. In this paper, we present an architecture that allows multiple clients protect the privacy of their own local data while the server only access target data. First, we analyze the effectiveness of using multiple sources, in domain adaptive object detection task. In client sides, we propose a source-only probabilistic teacher (PT) and leverage probabilistic teacher for domain adaptation (PTDA) as detectors to reduce false negatives. Moreover, we also introduce a Pseudo-label Voting Mechanism to filter out false positives with minimal communication costs. The performance of the proposed approach is evaluated on the ck2b and skf2c datasets and compared with other multi-source domain adaptation as well as federated learning methods. To sum up, the proposed method achieved better performance while preserving source data privacy and minimizing communication costs, without requiring the same model structure among different clients.

原文English
主出版物標題13th International Workshop on Computer Science and Engineering, WCSE 2023
發行者International Workshop on Computer Science and Engineering (WCSE)
頁面164-170
頁數7
ISBN(電子)9789811879500
DOIs
出版狀態Published - 2023
事件2023 13th International Workshop on Computer Science and Engineering, WCSE 2023 - Hybrid, Singapore, Singapore
持續時間: 16 6月 202318 6月 2023

出版系列

名字13th International Workshop on Computer Science and Engineering, WCSE 2023

Conference

Conference2023 13th International Workshop on Computer Science and Engineering, WCSE 2023
國家/地區Singapore
城市Hybrid, Singapore
期間16/06/2318/06/23

指紋

深入研究「Communication-efficient Multi-source Domain Adaptive Object Detection under Privacy Constraints」主題。共同形成了獨特的指紋。

引用此