Attention-Based Scattering Network for Satellite Imagery
This work addresses the need for more interpretable and data-efficient models for atmospheric forecasting, though it appears incremental as it builds on existing scattering transforms with added attention mechanisms.
The paper tackled the problem of effectively combining multi-channel satellite imagery features for atmospheric property prediction by introducing an attention-based scattering network that extracts high-level features without trainable parameters and uses a separation scheme for channel attention. It achieved promising results in estimating tropical cyclone intensity and predicting lightning occurrence from satellite imagery.
Multi-channel satellite imagery, from stacked spectral bands or spatiotemporal data, have meaningful representations for various atmospheric properties. Combining these features in an effective manner to create a performant and trustworthy model is of utmost importance to forecasters. Neural networks show promise, yet suffer from unintuitive computations, fusion of high-level features, and may be limited by the quantity of available data. In this work, we leverage the scattering transform to extract high-level features without additional trainable parameters and introduce a separation scheme to bring attention to independent input channels. Experiments show promising results on estimating tropical cyclone intensity and predicting the occurrence of lightning from satellite imagery.