Whistle variability and social acoustic interactions in bottlenose dolphins
summary
Semi-captive recordings from Dolphin Reef use a three-hydrophone array plus cameras and an energy-based contour extractor to pull 8,566 whistles (2019 subset) and later ~100 hours (2018–2024) processed with a CNN. Clustering and DTW reveal signature-whistle subcategories whose structured variability is neither emitter identity nor random noise. Markov-network analysis finds individual clusters, monologues from two leaders, a responding outsider, and nodes with high strength vs high betweenness suggesting hubs vs bridges; offspring produce deceased mothers’ signature whistles, implying displaced reference.
AI-written summary · about
chapters (11)
- 0:00Introduction and welcome
- 3:14Dolphin cognition and society
- 5:32Dolphin social behavior
- 8:47Data collection and signals
- 12:58Signature whistle analysis
- 16:47Whistle variability analysis
- 21:05Hypothesis on signal modulation
- 23:23Acoustic sequence networks
- 33:42Linguistic features in dolphins
- 38:16Future research directions
- 44:01Q&A and closing remarks
related talks
01h 01mIndividual Discrimination of Bottlenose Dolphin Signature Whistles Using Deep Learning
57mDecoding Dolphin Communication
01h 00mLong term work and Machine Learning Approaches to DolphinGemma
01h 07mDecoding Dolphin Speech
57mDo Elephants Have Names? Vocal Labeling of Individual Conspecifics in African Elephants + ...
25mThe Sound of Life
53mWhAM: Towards A Translative Model of Sperm Whale Vocalization
43mDeciphering animal emotions using machine learning and vocalisations