Common practices and taxonomy in deep multi-view fusion for remote sensing applications

• Code repository: github.com/fmenat/multiviewRS-models
• Datasets repository: github.com/fmenat/multiviewRS-datasets

This review provides a comprehensive overview and taxonomy of deep multi-view (MV) fusion for remote sensing and Earth Observation. It examines how multiple observations with heterogeneous characteristics—including different resolutions, temporal properties, sensor types, and noise levels—are integrated using neural network-based approaches. Given the inconsistent terminology and varying representations across the literature, the article proposes harmonized terminology and organizes common practices and concepts to establish a unified perspective on MV fusion. Focusing on supervised learning, the review synthesizes insights from a broad range of recent studies and provides an extensive reference base. It aims to support researchers in navigating the field and fostering more consistent future developments.

Recommended citation: Mena, Francisco, et al. "Common practices and taxonomy in deep multiview fusion for remote sensing applications." IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 17 (2024): 4797-4818.
Download Paper