URINet: Unsupervised point cloud rotation invariant representation learning via semantic and structural reasoning

Download PDF
Open Webpage

Qiuxia Wu, Kunming Su

Published: 01 Jan 2024, Last Modified: 05 Nov 2025Comput. Vis. Image Underst. 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•We proposed a novel two-branch encoder to address the information loss problem.•We reveal the self-reconstruction ambiguity and avoid with bidirectional alignment.•We present an unsupervised rotation invariant model to alleviate the label dependent.•We have achieved state of the art performance in downstream tasks.