Adversarially Trained Object Detector for Unsupervised Domain Adaptation

Adversarially Trained Object Detector for Unsupervised Domain Adaptation

Unsupervised domain adaptation, which involves transferring knowledge from a label-rich source domain to an unlabeled target domain, can be used to reduce annotation costs in the field of object detection substantially. This study demonstrates that adversarial training in the source domain can be em...

Full description

Bibliographic Details
Main Authors: Kazuma Fujii, Hiroshi Kera, Kazuhiko Kawamoto
Format: Article
Language:English
Published: IEEE 2022-01-01
Series:IEEE Access
Subjects:
Domain adaptation
object detection
adversarial training
Online Access:https://ieeexplore.ieee.org/document/9789122/

Internet

https://ieeexplore.ieee.org/document/9789122/

Similar Items