Abstract
Three-dimensional polygon meshes are among the most popular geometric data representations for many applications, which include three-dimensional object retrieval and classification. However, implementing a deep learning approach for three-dimensional data is a bit hard due to the complexity and irregularity of the mesh surface representation.In this paper, we propose a new geometric deep learning approach dedicated to representation learning, which applies convolutional operations on 3D shapes. In particular, we introduce a face-based convolutional operator that can learn highly discriminating features while avoiding high complexity and irregularity problems.We experimentally validated our approach on 3D shape classification and multi-domain protein shape retrieval challenge. A comparison with the state-of-the-art approaches proved the relevance of the learned features to classification accuracy and 3D object retrieval.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have