Self-training via Metric Learning for Source-Free Domain Adaptation of Semantic Segmentation

arXiv:2212.04227v2 Announce Type: replace-cross
Abstract: Unsupervised source-free domain adaptation methods aim to train a model for the target domain utilizing a pretrained source-domain model and unlabeled target-domain data, particularly when accessibility to source data is restricted due to intellectual property or privacy concerns. Traditional methods usually use self-training with pseudo-labeling, which is often subjected to thresholding based on prediction confidence. However, such thresholding limits the effectiveness of self-training due to insufficient supervision. This issue becomes more severe in a source-free …

abstract accessibility aim arxiv concerns cs.cv cs.lg data domain domain adaptation free intellectual property privacy property segmentation self-training semantic source data train training type unsupervised via