Segment scene in new target domain with unsupervised learning by paste target into source with EHTDI
Segment scene in new target domain with unsupervised learning by paste target into source with EHTDI
Exploring High-quality Target Domain Information for Unsupervised Domain Adaptive Semantic Segmentation
arXiv paper abstract https://arxiv.org/abs/2208.06100v1
arXiv PDF paper https://arxiv.org/pdf/2208.06100v1.pdf
In unsupervised domain adaptive (UDA) semantic segmentation ... distillation technique requires complicate multi-stage process and many training tricks.
... propose a simple yet effective ... idea is to fully explore the target-domain information from the views of boundaries and features.
... propose a novel mix-up strategy to generate high-quality target-domain boundaries with ground-truth labels.
... select the high-confidence target-domain areas and then paste them to the source-domain images.
... By combining two proposed methods, more discriminative features can be extracted and hard object boundaries can be better addressed for the target domain.
... for SYNTHIA -> Cityscapes ... state-of-the-art performance with 57.8% mIoU and 64.6% mIoU on 16 classes and 13 classes ...
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