Related Experiment Video
Updated: Jul 7, 2026
10:56
Long-term Behavioral Tracking of Freely Swimming Weakly Electric Fish
Published on: March 6, 2014
Automated whale shark recognition and tracking using drones and deep learning
Paola Judith Delgado-García1, Emiliano García-Rodríguez1, Oscar Sosa-Nishizaki1
1Department of Biological Oceanography, Centro de Investigación Científica y de Educación Superior de Ensenada, Baja California, Mexico.
Journal of Fish Biology
|July 5, 2026
Summary
Drone-based remote sensing combined with deep learning (DL) accurately identifies and tracks vulnerable whale sharks. This non-invasive method supports conservation efforts for marine megafauna.
Area of Science:
- Marine Biology
- Remote Sensing
- Artificial Intelligence
Background:
- Whale sharks (Rhincodon typus) are vulnerable migratory species forming aggregations in plankton-rich areas like Bahía de los Ángeles (BLA).
- Anthropogenic threats necessitate effective, non-invasive monitoring strategies for whale shark conservation.
Purpose of the Study:
- To explore the integration of drone remote sensing and deep learning (DL) for automated whale shark recognition and tracking.
- To evaluate the performance of different DL models in identifying whale sharks from drone imagery.
Main Methods:
- Conducted 59 drone flights in BLA, Mexico, resulting in 23 whale shark sightings.
- Implemented two DL approaches: DeepLabCut (pose estimation) and multi-scale patch (MSP) classification, utilizing convolutional neural networks, data augmentation, and transfer learning.
Main Results:
- The MSP approach with multi-size, overlapping patches achieved a high macro F1 score of 0.91.
- Environmental factors like turbidity and solar glare had minimal impact on the model's predictive performance.
- The DL-enhanced drone monitoring demonstrated scalable and accurate recognition and tracking capabilities.
Conclusions:
- Deep learning-enhanced drone monitoring offers a promising non-invasive tool for whale shark research and conservation.
- This technology can be applied in BLA and similar marine environments for effective ecological monitoring.
