The AIR Centre has announced an exciting new Kaggle competition: “Automatic Identification of Internal Waves: A Probabilistic Image Classification Challenge”. This competition, accessible on the AIR Centre’s Kaggle page, merges state-of-the-art satellite technology with advanced machine learning techniques.
Competition Highlights
- Objective: Develop an automated method to identify internal waves in satellite imagery.
- Dataset: Simplified Sentinel-1 (Copernicus Constellation) satellite images.
- Task: Image classification for internal wave detection.
- Deadline: October 31st, 2024
Why Focus on Oceanic Internal Waves?
Oceanic internal waves are unique waves that propagate within stratified ocean layers disturbed by physical mechanisms. These typically nonlinear waves, known as Internal Solitary Waves (ISWs), can exceed amplitudes of 100 meters. They generate the highest vertical velocities in the ocean and strong horizontal shear currents, often leading to underwater navigation accidents and damage to sea platforms. Additionally, ISWs resuspend sediments on the continental shelf and induce intense mixing in the deep ocean. ISWs can be detected by remote sensing satellites through variations in sea surface roughness, using optical and radar sensors. Understanding these complex waves is crucial for gaining insights into the intricate dynamics of the ocean.
This competition offers a unique opportunity to contribute to oceanographic research while honing machine learning skills. Whether the participants are a seasoned Kaggler or new to satellite imagery analysis, this challenge promises to be both educational and impactful.
Ready to make waves in the world of AI and ocean science? Head over to Kaggle now and dive into this fascinating competition!
