Image Labels Are All You Need for Coarse Seagrass Segmentation
Scarlett Raine, Ross Marchant, Brano Kusy, Frederic Maire, Tobias, Fischer

TL;DR
This paper introduces a weakly supervised approach for coarse seagrass segmentation using only image-level labels, leveraging large language models and contrastive learning to outperform previous patch-label methods.
Contribution
It proposes SeaFeats and SeaCLIP architectures that utilize unsupervised pre-training and language models, enabling effective seagrass classification with fewer labels and improved accuracy.
Findings
Outperforms previous patch-label methods by 6.8% in F1 score.
Achieves 12.1% higher F1 score on seagrass presence/absence detection.
Demonstrates robustness through real-world case studies.
Abstract
Seagrass meadows serve as critical carbon sinks, but estimating the amount of carbon they store requires knowledge of the seagrass species present. Underwater and surface vehicles equipped with machine learning algorithms can help to accurately estimate the composition and extent of seagrass meadows at scale. However, previous approaches for seagrass detection and classification have required supervision from patch-level labels. In this paper, we reframe seagrass classification as a weakly supervised coarse segmentation problem where image-level labels are used during training (25 times fewer labels compared to patch-level labeling) and patch-level outputs are obtained at inference time. To this end, we introduce SeaFeats, an architecture that uses unsupervised contrastive pre-training and feature similarity, and SeaCLIP, a model that showcases the effectiveness of large language models…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
Image Labels Are All You Need for Coarse Seagrass Segmentation· youtube
Taxonomy
TopicsMarine animal studies overview · Marine and coastal plant biology · Coral and Marine Ecosystems Studies
