Unsupervised Representation Learning for Diverse Deformable Shape Collections
This addresses the challenge of handling diverse 3D shapes in computer graphics and vision, offering a more flexible approach for applications like animation and modeling, though it is incremental in improving unsupervised representation learning.
The paper tackles the problem of encoding and manipulating 3D surface meshes for deformable shape collections by introducing an unsupervised learning method that creates an interpretable embedding space without requiring meshes to be in 1-to-1 correspondence, achieving excellent reconstructions and smoother interpolations than baselines.
We introduce a novel learning-based method for encoding and manipulating 3D surface meshes. Our method is specifically designed to create an interpretable embedding space for deformable shape collections. Unlike previous 3D mesh autoencoders that require meshes to be in a 1-to-1 correspondence, our approach is trained on diverse meshes in an unsupervised manner. Central to our method is a spectral pooling technique that establishes a universal latent space, breaking free from traditional constraints of mesh connectivity and shape categories. The entire process consists of two stages. In the first stage, we employ the functional map paradigm to extract point-to-point (p2p) maps between a collection of shapes in an unsupervised manner. These p2p maps are then utilized to construct a common latent space, which ensures straightforward interpretation and independence from mesh connectivity and shape category. Through extensive experiments, we demonstrate that our method achieves excellent reconstructions and produces more realistic and smoother interpolations than baseline approaches.