Can I read Unsupervised 3D Shape Co-Segmentation Techniques on EtoBox?

Unsupervised 3D Shape Co-Segmentation Techniques by abouqora is a document available to read on EtoBox.

What is Unsupervised 3D Shape Co-Segmentation Techniques about?

This document presents an unsupervised approach for 3D shape co-segmentation using local geometric features, focusing on segmenting unlabeled 3D shapes into distinct parts and establishing correspondences between them. The method employs a Stacked Sparse AutoEncoder for representation learning and integrates a Gaussian Mixture Model and Conditional Random Fields to enhance label consistency and refine boundaries. Experimental results demonstrate improved co-segmentation outcomes, validated using the Princet

Author

abouqora

Language

EN