A Multi-Scale Attention-Enhanced Architecture for Gravity Wave Localization in Satellite Imagery

TL;DR

This paper introduces YOLO-DCAT, an advanced neural network architecture that enhances gravity wave detection in satellite images by effectively capturing multi-scale features and focusing on relevant regions, significantly improving localization accuracy.

Contribution

The paper presents a novel YOLO-DCAT model with Multi Dilated Residual Convolution and Simplified Spatial and Channel Attention, tailored for complex gravity wave localization in satellite imagery.

Findings

Over 14% increase in mean Average Precision (mAP)

Approximately 17% improvement in Intersection over Union (IoU)

Significant enhancement in localization accuracy for gravity waves

Abstract

Satellite images present unique challenges due to their high object variability and lower spatial resolution, particularly for detecting atmospheric gravity waves which exhibit significant variability in scale, shape, and pattern extent, making accurate localization highly challenging. This variability is further compounded by dominant unwanted objects such as clouds and city lights, as well as instrumental noise, all within a single image channel, while conventional detection methods struggle to capture the diverse and often subtle features of gravity waves across varying conditions. To address these issues, we introduce YOLO-DCAT incorporating Multi Dilated Residual Convolution (MDRC) and Simplified Spatial and Channel Attention (SSCA), an enhanced version of YOLOv5 specifically designed to improve gravity wave localization by effectively handling their complex and variable characteristics. MDRC captures multi-scale features through parallel dilated convolutions with varying dilation rates, while SSCA focuses on the most relevant spatial regions and channel features to enhance detection accuracy and suppress interference from background noise. In our experiments, the improved model outperformed state-of-the-art alternatives, improving mean Average Precision (mAP) by over 14% and Intersection over Union (IoU) by approximately 17%, demonstrating significantly improved localization accuracy for gravity waves in challenging satellite imagery and contributing to more precise climate research and modeling.