SAL: Sign Agnostic Learning of Shapes from Raw Data
This method addresses a bottleneck in geometric deep learning by enabling direct use of real-world data, reducing manual preprocessing for applications like surface reconstruction and shape space learning.
The paper tackles the problem of learning implicit shape representations without requiring precomputed signed distance or occupancy functions, achieving state-of-the-art reconstructions from raw, unsigned data like point clouds and triangle soups.
Recently, neural networks have been used as implicit representations for surface reconstruction, modelling, learning, and generation. So far, training neural networks to be implicit representations of surfaces required training data sampled from a ground-truth signed implicit functions such as signed distance or occupancy functions, which are notoriously hard to compute. In this paper we introduce Sign Agnostic Learning (SAL), a deep learning approach for learning implicit shape representations directly from raw, unsigned geometric data, such as point clouds and triangle soups. We have tested SAL on the challenging problem of surface reconstruction from an un-oriented point cloud, as well as end-to-end human shape space learning directly from raw scans dataset, and achieved state of the art reconstructions compared to current approaches. We believe SAL opens the door to many geometric deep learning applications with real-world data, alleviating the usual painstaking, often manual pre-process.