Dynamic mapping from static labels: remote sensing dynamic sample generation with temporal-spectral embedding

TL;DR

This paper introduces TasGen, a novel two-stage method that generates dynamic training samples from static labels in remote sensing by disentangling and modeling temporal and spectral features, enabling better detection and explanation of land surface changes.

Contribution

TasGen is the first approach to generate dynamic samples from static labels using a hierarchical temporal-spectral autoencoder and anomaly interpretation, addressing land surface change detection without manual updates.

Findings

Effective anomaly detection via joint temporal-spectral embedding

Automatic relabeling of change points for dynamic sample generation

Interpretation of land surface dynamics through spectral-temporal attribution

Abstract

Accurate remote sensing geographic mapping requires timely and representative samples. However, rapid land surface changes often render static samples obsolete within months, making manual sample updates labor-intensive and unsustainable. To address this challenge, we propose TasGen, a two-stage Temporal spectral-aware Automatic Sample Generation method for generating dynamic training samples from single-date static labels without human intervention. Land surface dynamics often manifest as anomalies in temporal-spectral sequences. %These anomalies are multivariate yet unified: temporal, spectral, or joint anomalies stem from different mechanisms and cannot be naively coupled, as this may obscure the nature of changes. Yet, any land surface state corresponds to a coherent temporal-spectral signature, which would be lost if the two dimensions are modeled separately. To effectively capture these dynamics, TasGen first disentangles temporal and spectral features to isolate their individual contributions, and then couples them to model their synergistic interactions. In the first stage, we introduce a hierarchical temporal-spectral variational autoencoder (HTS-VAE) with a dual-dimension embedding to learn low-dimensional latent patterns of normal samples by first disentangling and then jointly embedding temporal and spectral information. This temporal-spectral embedding enables robust anomaly detection by identifying deviations from learned joint patterns. In the second stage, a classifier trained on stable samples relabels change points across time to generate dynamic samples. To not only detect but also explain surface dynamics, we further propose an anomaly interpretation method based on Gibbs sampling, which attributes changes to specific spectral-temporal dimensions.