Gappy local conformal auto-encoders for heterogeneous data fusion: in praise of rigidity
This addresses the challenge of integrating diverse sensor data for applications like localization and PDE analysis, but it appears incremental as it builds on existing auto-encoder methods for data fusion.
The paper tackles the problem of fusing heterogeneous, partial measurements from multiple sensors by proposing an end-to-end computational pipeline using a multiple-auto-encoder neural network architecture, resulting in a globally consistent latent space that harmonizes all measurements. It demonstrates the approach on examples including a Wi-Fi localization problem and a spatio-temporal dynamical puzzle, though no concrete numerical results are provided.
Fusing measurements from multiple, heterogeneous, partial sources, observing a common object or process, poses challenges due to the increasing availability of numbers and types of sensors. In this work we propose, implement and validate an end-to-end computational pipeline in the form of a multiple-auto-encoder neural network architecture for this task. The inputs to the pipeline are several sets of partial observations, and the result is a globally consistent latent space, harmonizing (rigidifying, fusing) all measurements. The key enabler is the availability of multiple slightly perturbed measurements of each instance:, local measurement, "bursts", that allows us to estimate the local distortion induced by each instrument. We demonstrate the approach in a sequence of examples, starting with simple two-dimensional data sets and proceeding to a Wi-Fi localization problem and to the solution of a "dynamical puzzle" arising in spatio-temporal observations of the solutions of Partial Differential Equations.