A Latent Representation Learning Framework for Hyperspectral Image Emulation in Remote Sensing

TL;DR

This paper introduces a latent representation framework for generating synthetic hyperspectral images that improves accuracy and robustness over traditional models, supporting both spectral and spatial emulation for remote sensing.

Contribution

It presents a novel latent representation-based approach for hyperspectral emulation, combining VAE pretraining with parameter interpolation, advancing the state-of-the-art in synthetic hyperspectral data generation.

Findings

Outperforms classical regression-based emulators in accuracy and spectral fidelity

Demonstrates robustness to real-world spatial variability

Preserves downstream biophysical parameter retrieval performance

Abstract

Synthetic hyperspectral image (HSI) generation is essential for large-scale simulation, algorithm development, and mission design, yet traditional radiative transfer models remain computationally expensive and often limited to spectrum-level outputs. In this work, we propose a latent representation-based framework for hyperspectral emulation that learns a latent generative representation of hyperspectral data. The proposed approach supports both spectrum-level and spatial-spectral emulation and can be trained either in a direct one-step formulation or in a two-step strategy that couples variational autoencoder (VAE) pretraining with parameter-to-latent interpolation. Experiments on PROSAIL-simulated vegetation data and Sentinel-3 OLCI imagery demonstrate that the method outperforms classical regression-based emulators in reconstruction accuracy, spectral fidelity, and robustness to real-world spatial variability. We further show that emulated HSIs preserve performance in downstream biophysical parameter retrieval, highlighting the practical relevance of emulated data for remote sensing applications.