Geometric Autoencoders -- What You See is What You Decode

2023/06/30 by Philipp Nazari, Nazari, Philipp, Sebastian Damrich +3 · 6 citations
Computer Science · #AI in cancer detection #Data Analysis with R #Data Visualization and Analytics #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2306.17638

openalex publication_date 2023/06/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/01

Abstract

Visualization is a crucial step in exploratory data analysis. One possible approach is to train an autoencoder with low-dimensional latent space. Large network depth and width can help unfolding the data. However, such expressive networks can achieve low reconstruction error even when the latent representation is distorted. To avoid such misleading visualizations, we propose first a differential geometric perspective on the decoder, leading to insightful diagnostics for an embedding's distortion, and second a new regularizer mitigating such distortion. Our ``Geometric Autoencoder'' avoids stretching the embedding spuriously, so that the visualization captures the data structure more faithfully. It also flags areas where little distortion could not be achieved, thus guarding against misinterpretation.

Cited by

• Multivariate Time Series Forecasting with Hybrid Euclidean-SPD Manifold Graph Neural Networks
• Multi-Scale Geometric Autoencoder
• Geometric Autoencoder Priors for Bayesian Inversion: Learn First Observe Later
• DREAMS: Preserving both Local and Global Structure in Dimensionality Reduction
• Enforcing Latent Euclidean Geometry in Single-Cell VAEs for Manifold Interpolation
• Nonlinear Model Order Reduction of Dynamical Systems in Process Engineering: Review and Comparison

Related

• Multi-Scale Geometric Autoencoder
• A Geometric Perspective on Autoencoders
• Processsing Simple Geometric Attributes with Autoencoders
• Isometric Autoencoders
• A PCA-like Autoencoder