Beluga Whale Detection from Satellite Imagery with Point Labels
Yijie Zheng, Jinxuan Yang, Yu Chen, Yaxuan Wang, Yihang Lu, Guoqing Li

TL;DR
This paper presents an automated, point-label-based detection pipeline for beluga whales and seals in satellite imagery, improving accuracy and reducing manual annotation effort using SAM and YOLOv8.
Contribution
It introduces a novel detection pipeline that leverages point annotations and SAM to generate bounding boxes, enhancing detection of uncertain whales and streamlining the annotation process.
Findings
SAM-generated annotations outperform traditional buffer-based methods
YOLOv8 achieves 72.2% F1-score for whales
Method improves detection in dense scenes
Abstract
Very high-resolution (VHR) satellite imagery has emerged as a powerful tool for monitoring marine animals on a large scale. However, existing deep learning-based whale detection methods usually require manually created, high-quality bounding box annotations, which are labor-intensive to produce. Moreover, existing studies often exclude ``uncertain whales'', individuals that have ambiguous appearances in satellite imagery, limiting the applicability of these models in real-world scenarios. To address these limitations, this study introduces an automated pipeline for detecting beluga whales and harp seals in VHR satellite imagery. The pipeline leverages point annotations and the Segment Anything Model (SAM) to generate precise bounding box annotations, which are used to train YOLOv8 for multiclass detection of certain whales, uncertain whales, and harp seals. Experimental results…
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
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMarine animal studies overview · Advanced Neural Network Applications · Infrared Target Detection Methodologies
MethodsYou Only Look Once
