Self-supervised Learning of Dense Shape Correspondence

Abstract

We introduce the first completely unsupervised correspondence learning approach for deformable 3D shapes. Key to our model is the understanding that natural deformations (such as changes in pose) approximately preserve the metric structure of the surface, yielding a natural criterion to drive the learning process toward distortion-minimizing predictions. On this basis, we overcome the need for annotated data and replace it by a purely geometric criterion. The resulting learning model is class-agnostic, and is able to leverage any type of deformable geometric data for the training phase. In contrast to existing supervised approaches which specialize on the class seen at training time, we demonstrate stronger generalization as well as applicability to a variety of challenging settings. We showcase our method on a wide selection of correspondence benchmarks, where we outperform other methods in terms of accuracy, generalization, and efficiency.

Cited by in corpus (8)

• Dynamic Graph CNN for Learning on Point Clouds
• Deep Graph Matching Consensus
• Semantic Correspondence via 2D-3D-2D Cycle
• Continuous Geodesic Convolutions for Learning on 3D Shapes
• Data Transformation Insights in Self-supervision with Clustering Tasks
• Human Correspondence Consensus for 3D Object Semantic Understanding
• Unsupervised Scale-Invariant Multispectral Shape Matching
• Efficient and Robust Shape Correspondence via Sparsity-Enforced Quadratic Assignment