Attention-Based Scattering Network for Satellite Imagery

TL;DR

This paper introduces an attention-based scattering network that effectively fuses multi-channel satellite imagery features for atmospheric property estimation, improving interpretability and performance without requiring extensive training data.

Contribution

It proposes a novel scattering transform-based neural network with an attention mechanism for channel separation, enhancing feature extraction and fusion in satellite imagery analysis.

Findings

Improved tropical cyclone intensity estimation

Enhanced lightning occurrence prediction

Effective feature extraction without additional trainable parameters

Abstract

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.