What question did this study set out to answer?

This study aims to classify dolphin whistle producers, extending beyond just signature whistles to include other types.

June 28, 2026Marine Mammal ScienceOpen Access

Identifying Dolphin Whistle Producers With Deep Learning: Moving Beyond Signature Whistles

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial evaluates classification of dolphin whistles using deep learning, indicating potential for broader identification methods.

This study aims to classify dolphin whistle producers, extending beyond just signature whistles to include other types.
Utilized a convolutional neural network to analyze raw amplitude vectors from whistle recordings.
Classified whistles from 20 different dolphins, regardless of contour type.
Evaluated model performance across various whistle categories including signature, shared, variant, and copies.
Achieved overall classification accuracy of 85.20%; 96.10% for signature whistles, 92.50% for shared whistles, 72.10% for variant whistles, and 17.20% for copies.
Model trained solely on signature whistles reached 93.70% accuracy, enhancing trade-off evaluation between ecological relevance and accuracy.

Jones et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

June 28, 2026Marine Mammal ScienceOpen Access

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial evaluates classification of dolphin whistles using deep learning, indicating potential for broader identification methods.

This study aims to classify dolphin whistle producers, extending beyond just signature whistles to include other types.
Utilized a convolutional neural network to analyze raw amplitude vectors from whistle recordings.
Classified whistles from 20 different dolphins, regardless of contour type.
Evaluated model performance across various whistle categories including signature, shared, variant, and copies.
Achieved overall classification accuracy of 85.20%; 96.10% for signature whistles, 92.50% for shared whistles, 72.10% for variant whistles, and 17.20% for copies.
Model trained solely on signature whistles reached 93.70% accuracy, enhancing trade-off evaluation between ecological relevance and accuracy.

Jones et al. (2026) studied this question.

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark