COP-GEN: Latent Diffusion Transformer for Copernicus Earth Observation Data -- Generation Stochastic by Design
This addresses the need for stochastic generative modeling in Earth observation applications to handle non-injective relationships between sensors, improving data integration and uncertainty representation for tasks like data completion.
The paper tackles the problem of generating diverse and physically plausible Earth observation data from multiple sensors by introducing COP-GEN, a latent diffusion transformer that models conditional distributions, enabling tasks like zero-shot modality translation and spectral band infilling without retraining, with experiments showing strong peak fidelity and meaningful cross-modal structure.
Earth observation applications increasingly rely on data from multiple sensors, including optical, radar, elevation, and land-cover products. Relationships between these modalities are fundamental for data integration but are inherently non-injective: identical conditioning information can correspond to multiple physically plausible observations. Thus, such conditional mappings should be parametrised as data distributions. As a result, deterministic models tend to collapse toward conditional means and fail to represent the uncertainty and variability required for tasks such as data completion and cross-sensor translation. We introduce COP-GEN, a multimodal latent diffusion transformer that models the joint distribution of heterogeneous Earth Observation modalities at their native spatial resolutions. By parameterising cross-modal mappings as conditional distributions, COP-GEN enables flexible any-to-any conditional generation, including zero-shot modality translation, spectral band infilling, and generation under partial or missing inputs, without task-specific retraining. Experiments on a large-scale global multimodal dataset show that COP-GEN generates diverse yet physically consistent realisations while maintaining strong peak fidelity across optical, radar, and elevation modalities. Qualitative and quantitative analyses demonstrate that the model captures meaningful cross-modal structure and systematically adapts its output uncertainty as conditioning information increases. These results highlight the practical importance of stochastic generative modeling for Earth observation and motivate evaluation protocols that move beyond single-reference, pointwise metrics. Website: https:// miquel-espinosa.github.io/cop-gen