Latent Space Inference via Paired Autoencoders

2026-01-16

AI Summary

This work proposes a data-driven latent-space inference framework based on paired autoencoders to address parameter estimation bias in inverse problems caused by missing, noisy, or out-of-distribution observations. The approach constructs separate autoencoders in the parameter and observation spaces and learns a mapping between their latent representations, enabling low-dimensional, information-preserving regularized inversion. By jointly optimizing data reconstruction and parameter estimation, the method significantly enhances reconstruction accuracy while maintaining physical consistency. Evaluated on medical tomographic imaging and geophysical full-waveform inversion tasks, the proposed framework demonstrates superior robustness and accuracy compared to conventional end-to-end encoder-decoder architectures and single-branch autoencoder baselines.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Intelligent Robots: State Estimation
• Reasoning under Uncertainty: Probabilistic Inference

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

This work describes a novel data-driven latent space inference framework built on paired autoencoders to handle observational inconsistencies when solving inverse problems. Our approach uses two autoencoders, one for the parameter space and one for the observation space, connected by learned mappings between the autoencoders'latent spaces. These mappings enable a surrogate for regularized inversion and optimization in low-dimensional, informative latent spaces. Our flexible framework can work with partial, noisy, or out-of-distribution data, all while maintaining consistency with the underlying physical models. The paired autoencoders enable reconstruction of corrupted data, and then use the reconstructed data for parameter estimation, which produces more accurate reconstructions compared to paired autoencoders alone and end-to-end encoder-decoders of the same architecture, especially in scenarios with data inconsistencies. We demonstrate our approaches on two imaging examples in medical tomography and geophysical seismic-waveform inversion, but the described approaches are broadly applicable to a variety of inverse problems in scientific and engineering applications.

Problem

• inverse problems
• observational inconsistencies
• latent space inference
• data reconstruction
• parameter estimation

Innovation

• paired autoencoders
• latent space inference
• inverse problems
• data reconstruction
• regularized inversion