CNN 모델을 이용한 수중음향 자료의 돌고래 휘슬음 자동 분류 연구
Research on Automatic Classification of Dolphin Whistle in Underwater Acoustic Data Using Convolution Neural Network (CNN) Model
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
With the advancement of underwater acoustic observation technology and the development of high-performance equipment, passive acoustic monitoring (PAM) has become a widely used method for monitoring ecosystem cetaceans. As the amount of high-resolution acoustic data increases exponentially, researchers continuously strive to automate the classification of marine mammal sounds and improve the speed of acoustic data analysis. This study developed a method for automatically classifying the dolphin whistles of common dolphins, false killer whales, and bottlenose dolphins using a convolutional neural network model. Whistle signals were analyzed using three spectrogram images: Original, median filter, and edge detection. The F1-Scores analysis results for the three data types were 84.40%, 84.65%, and 80.29%, confirming the possibility of automating whistle classification for the three dolphin species. In Case 2 of the F1-Scores, the common dolphin scored 90.90%, the false killer whale scored 80.00%, and the bottlenose dolphin scored 85.71%, indicating that the common dolphin showed relatively high automatic detection. This study contributes to developing an automatic detection and classification technique for cetacean species in South Korea. Long-term underwater acoustic measurement data acquired through PAM can be utilized more effectively in the future.
Abstract
With the advancement of underwater acoustic observation technology and the development of high-performance equipment, passive acoustic monitoring (PAM) has become a widely used method for monitoring ecosystem cetaceans. As the amount of high-resolution acoustic data increases exponentially, researchers continuously strive to automate the classification of marine mammal sounds and improve the speed of acoustic data analysis. This study developed a method for automatically classifying the dolphin whistles of common dolphins, false killer whales, and bottlenose dolphins using a convolutional neural network model. Whistle signals were analyzed using three spectrogram images: Original, median filter, and edge detection. The F1-Scores analysis results for the three data types were 84.40%, 84.65%, and 80.29%, confirming the possibility of automating whistle classification for the three dolphin species. In Case 2 of the F1-Scores, the common dolphin scored 90.90%, the false killer whale scored 80.00%, and the bottlenose dolphin scored 85.71%, indicating that the common dolphin showed relatively high automatic detection. This study contributes to developing an automatic detection and classification technique for cetacean species in South Korea. Long-term underwater acoustic measurement data acquired through PAM can be utilized more effectively in the future.
참고문헌 (46)
- Allen AN, "A convolutional neural network for automated detection of humpback whale song in a diverse, long-term passive acoustic dataset", Front Mar Sci, 8, pp.607321~ (2021)[DOI]
- André M, "Listening to the deep : Live monitoring of ocean noise and cetacean acoustic signals", Mar Pollut Bull, 63, pp.18~26 (2011)[DOI]
- Antichi S, "Differences in the whistles of two ecotypes of bottlenose dolphins from the Gulf of California", JASA Express Lett, 3, pp.051201~ (2023)[DOI]
- Bergler C, "ANIMAL-SPOT enables animal-independent signal detection and classification using deep learning", Sci Rep, 12, pp.21966~ (2022)[DOI]
- Bittle M, "A review of current marine mammal detection and classification algorithms for use in automated passive acoustic monitoring", Proceedings of Acoustics, pp.8~ (2013)
- Burtenshaw JC, "Acoustic and satellite remote sensing of blue whale seasonality and habitat in the Northeast Pacific", Deep Sea Res II Top Stud Oceanogr, 51, pp.967~986 (2004)[DOI]
- Carlón-Beltrán Ó, "Whistle characterization of longbeaked common dolphin(Delphinus delphis bairdii)in La Paz Bay, Gulf of California", PeerJ, 11, pp.e15687~ (2023)[DOI]
- Caruso F, "Monitoring of a nearshore small dolphin species using passive acoustic platforms and supervised machine learning techniques", Front Mar Sci, 7, pp.267~ (2020)[DOI]
- Choi KH, "Analysis of echolocation click signals of Indo-Pacific bottlenose dolphin(Tursiops aduncus)in Jeju Island", J Acoust Soc Kr, 38, pp.56~65 (2019)[DOI]
- Choi S, "Whistle characteristics of captive beluga whale Delphinapterus leucas", Korean J Fish Aquat Sci, 50, pp.799~805 (2017)[DOI]
- Clark CW, "Vocal activity of fin whales, Balaenoptera physalus, in the Ligurian Sea", Mar Mamm Sci, 18, pp.286~295 (2002)[DOI]
- Dosovitskiy A, "An image is worth 16x16 words : Transformers for image recognition at scale", ICLR 2020, pp.1~22 (2020)
- Frainer G, "Automatic detection and taxonomic identification of dolphin vocalisations using convolutional neural networks for passive acoustic monitoring", Eco Inform, 78, pp.102291~ (2023)[DOI]
- Gibb R, "Emerging opportunities and challenges for passive acoustics in ecological assessment and monitoring", Methods Ecol Evol, 10, pp.169~185 (2019)[DOI]
- "Metrics for multiclass classification: An overview" (2020)[URL]
- Gridley T, "Towed passive acoustic monitoring complements visual survey methods for Heaviside’s dolphins Cephalorhynchus heavisidii in the Namibian Islands Marine Protected Area", Afr J Mar Sci, 42, pp.495~506 (2020)[DOI]
- Hatamizadeh A, "Global context vision transformers", Proceedings of the 40th International Conference on Machine Learning. PMLR 202, pp.12633~12646 (2023)
- Huang G, "Densely connected convolutional networks", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.4700~4708 (2017)
- Jeong S, "Development of dolphin click signal classification algorithm based on recurrent neural network for marine environment monitoring", Geophys Geophys Explor, 26, pp.126~137 (2023)[DOI]
- Kim H, "Occurrence and acoustic characteristics of finless porpoise (Neophocaena asiaeorientalis) off the south coast of Korea using hydrophone and visual surveys", J Korean Soc Fish Ocean Technol, 60, pp.258~268 (2024)[DOI]
- Kim JS, "Classification of bearded seals signal based on convolutional neural network", J Acoust Soc Kr, 41, pp.235~241 (2022)[DOI]
- Kim S, "Application of the artificial intelligence for automatic detection of shipping noise in shallow-water", J Acoust Soc Kr, 39, pp.279~285 (2020)[DOI]
- Küsel ET, "Single-sensor, cue-counting population density estimation : Average probability of detection of broadband clicks", J Acoust Soc Am, 140, pp.1894~1903 (2016)[DOI]
- Lee JH, "Occurrence and spatial distribution of marine mammals by sighting surveys in Korean waters during 2011-2020", Korean J Fish Aquat Sci, 55, pp.938~945 (2022)[DOI]
- Lindseth A, "Underwater soundscape monitoring and fish bioacoustics : A review", Fishes, 3, pp.36~ (2018)[DOI]
- Li Y, "Computer Vision – ECCV 2022" (2022), Springer
- Marques TA, "Estimating animal population density using passive acoustics", Biol Rev, 88, pp.287~309 (2013)[DOI]
- McCullough JLK, "Acoustic classification of false killer whales in the Hawaiian Islands based on comprehensive vocal repertoire", JASA Express Lett, 1, pp.071201~ (2021)[DOI]
- McDonald MA, "Passive acoustic methods applied to fin whale population density estimation", J Acoust Soc Am, 105, pp.2643~2651 (1999)[DOI]
- Mellinger DK, "An overview of fixed passive acoustic observation methods for cetaceans", Oceanography, 20, pp.36~45 (2007)[DOI]
- Munger LM, "North Pacific right whale(Eubalaena japonica)seasonal and diel calling patterns from long-term acoustic recordings in the southeastern Bering Sea, 2000-2006", Mar Mamm Sci, 24, pp.795~814 (2008)[DOI]
- Oleson EM, "Behavioral context of call production by eastern North Pacific blue whales", Mar Ecol Prog Ser, 330, pp.269~284 (2007)[DOI]
- Parsons ECM, "Key research questions of global importance for cetacean conservation", Endanger Species Res, 27, pp.113~118 (2015)[DOI]
- Putland RL, "A hidden Markov model approach to indicate Bryde’s whale acoustics", Ecol Indic, 84, pp.479~487 (2018)[DOI]
- Shamir L, "Classification of large acoustic datasets using machine learning and crowdsourcing : Application to whale calls", J Acoust Soc Am, 135, pp.953~962 (2014)[DOI]
- Shin HL, "Characteristics of bottlenose dolphin(Tursiops truncatus)whistle", J Kor Soc Fish Technol, 38, pp.271~277 (2002)
- Širović A, "Seasonality of blue and fin whale calls west of the Antarctic Peninsula", Proceedings of the Oceans 2003, pp.818~ (2003)
- Stafford KM, "An acoustic link between blue whales in the eastern tropical Pacific and the Northeast Pacific", Mar Mamm Sci, 15, pp.1258~1268 (1999)[DOI]
- Usman AM, "Review of automatic detection and classification techniques for cetacean vocalization", IEEE Access, 8, pp.105181~105206 (2020)[DOI]
- Weinstein BG, "A computer vision for animal ecology", J Anim Ecol, 87, pp.533~545 (2018)[DOI]
- White EL, "More than a whistle : Automated detection of marine sound sources with a convolutional neural network", Front Mar Sci, 9, pp.879145~ (2022)[DOI]
- Winn HE, "The song of the humpback whale(Megaptera novaeangliae)in the West Indies", Mar Biol, 47, pp.97~114 (1978)[DOI]
- Yoon YG, "Study of acoustic characteristics of common dolphins Delphinus delphis in the East Sea", Korean J Fish Aquat Sci, 50, pp.406~412 (2017)[DOI]
- Yoon YG, "Preliminary study of passive acoustic monitoring of finless porpoises Neophocaena asiaeorientalis around the Southwest offshore wind farm in Korea", J Acoust Soc Kr, 40, pp.537~545 (2021)[DOI]
- Ziegenhorn MA, "Discriminating and classifying odontocete echolocation clicks in the Hawaiian Islands using machine learning methods", PLoS One, 17, pp.e0266424~ (2022)[DOI]
- Zimmer WMX, "Passive acoustic monitoring of cetaceans" (2011), Cambridge University Press
