Contrastive ground-level image and remote sensing pre-training improves representation learning for natural world imagery

TL;DR

This paper introduces CRISP, a contrastive pre-training method utilizing ground-level and aerial images, which enhances species recognition performance by leveraging multiple views in natural world imagery.

Contribution

The paper presents a novel contrastive pre-training task for ground-level and aerial images, along with a large multi-view dataset for ecological species recognition.

Findings

Improved fine-grained classification accuracy for species recognition.

Effective use of multi-view contrastive learning with limited view availability.

Introduction of a new large-scale natural world imagery dataset.

Abstract

Multimodal image-text contrastive learning has shown that joint representations can be learned across modalities. Here, we show how leveraging multiple views of image data with contrastive learning can improve downstream fine-grained classification performance for species recognition, even when one view is absent. We propose ContRastive Image-remote Sensing Pre-training (CRISP)$\unicode{x2014}$a new pre-training task for ground-level and aerial image representation learning of the natural world$\unicode{x2014}$and introduce Nature Multi-View (NMV), a dataset of natural world imagery including $>3$ million ground-level and aerial image pairs for over 6,000 plant taxa across the ecologically diverse state of California. The NMV dataset and accompanying material are available at hf.co/datasets/andyvhuynh/NatureMultiView.