Segment scene in new domain by reduce bias when mix source and target domain with Guidance-Training
Segment scene in new domain by reduce bias when mix source and target domain with Guidance-Training
Improve Cross-domain Mixed Sampling with Guidance Training for Adaptive Segmentation
arXiv paper abstract https://arxiv.org/abs/2403.14995
arXiv PDF paper https://arxiv.org/pdf/2403.14995.pdf
Unsupervised Domain Adaptation (UDA) … adjust models trained on a source domain to perform … on a target domain without … additional annotations … domain adaptive semantic segmentation … tackles UDA for dense prediction … goal is to circumvent the need for … annotations.
… propose a novel auxiliary task called Guidance Training.
This task facilitates the effective utilization of cross-domain mixed sampling techniques while mitigating distribution shifts from the real world.
… Guidance Training guides the model to extract and reconstruct the target-domain feature distribution from mixed data, followed by decoding the reconstructed target-domain features to make pseudo-label predictions.
Importantly, integrating Guidance Training incurs minimal training overhead and imposes no additional inference burden.
… demonstrate the efficacy of … approach by integrating it with existing methods, consistently improving performance …
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
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
