Detect object with DETR in a new domain unsupervised by randomly mask features with MTM
Detect object with DETR in a new domain unsupervised by randomly mask features with MTM
Mean Teacher DETR with Masked Feature Alignment: A Robust Domain Adaptive Detection Transformer Framework
arXiv paper abstract https://arxiv.org/abs/2310.15646
arXiv PDF paper https://arxiv.org/pdf/2310.15646.pdf
Unsupervised domain adaptation object detection(UDAOD) research on Detection Transformer(DETR) mainly focuses on feature alignment and existing methods
... propose a two-stage framework named MTM ... Mean Teacher-DETR with Masked Feature Alignment.
In the pretraining stage, ... utilize labeled target-like images produced by image style transfer to avoid performance fluctuation.
In the self-training stage, ... leverage unlabeled target images by pseudo labels based on mean teacher and propose a module called Object Queries Knowledge Transfer(OQKT) to ensure consistent performance gains of the student model.
... propose masked feature alignment methods including Masked Domain Query-based Feature Alignment(MDQFA) and Masked Token-wise Feature Alignment(MTWFA) to alleviate domain shift ... prevent training stagnation and lead to a robust pretrained model in the pretraining stage ... enhance the model's target performance in the self-training stage.
... verify the effectiveness of MTM.
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments