360SFUDA++: Towards Source-free UDA for Panoramic Segmentation by Learning Reliable Category Prototypes

arXiv:2404.16501v1 Announce Type: new
Abstract: In this paper, we address the challenging source-free unsupervised domain adaptation (SFUDA) for pinhole-to-panoramic semantic segmentation, given only a pinhole image pre-trained model (i.e., source) and unlabeled panoramic images (i.e., target). Tackling this problem is non-trivial due to three critical challenges: 1) semantic mismatches from the distinct Field-of-View (FoV) between domains, 2) style discrepancies inherent in the UDA problem, and 3) inevitable distortion of the panoramic images. To tackle these problems, we propose 360SFUDA++ that …

abstract arxiv challenges cs.cv domain domain adaptation free image images paper pre-trained model segmentation semantic type unsupervised