Tile Compression and Embeddings for Multi-Label Classification in GeoLifeCLEF 2024

TL;DR

This paper presents a multi-faceted approach combining frequency-domain compression, neighborhood models, and self-supervised learning to improve plant species classification from remote sensing data in the GeoLifeCLEF 2024 competition.

Contribution

It introduces the use of DCT-based data compression and LSH-based neighborhood models, along with tile2vec embeddings, for enhanced multi-label classification in geospatial data.

Findings

Best model achieved a leaderboard score of 0.152

Post-competition score improved to 0.161

Source code and models are publicly available

Abstract

We explore methods to solve the multi-label classification task posed by the GeoLifeCLEF 2024 competition with the DS@GT team, which aims to predict the presence and absence of plant species at specific locations using spatial and temporal remote sensing data. Our approach uses frequency-domain coefficients via the Discrete Cosine Transform (DCT) to compress and pre-compute the raw input data for convolutional neural networks. We also investigate nearest neighborhood models via locality-sensitive hashing (LSH) for prediction and to aid in the self-supervised contrastive learning of embeddings through tile2vec. Our best competition model utilized geolocation features with a leaderboard score of 0.152 and a best post-competition score of 0.161. Source code and models are available at https://github.com/dsgt-kaggle-clef/geolifeclef-2024.