Methods to Improve the Robustness of Right Whale Detection using CNNs in Changing Conditions. Vickers, W., Milner, B., Gorpincenko, A., & Lee, R. In 2020 28th European Signal Processing Conference (EUSIPCO), pages 106-110, Aug, 2020. Paper doi abstract bibtex This paper is concerned with developing a method of detecting right whales from autonomous surface vehicles (ASVs) that is robust to changing operating conditions. A baseline convolutional neural network (CNN) is first trained using data taken from a single operating condition. Its detection accuracy is then found to degrade when applied to different operating conditions. Two methods are then investigated to restore performance using just a single model. The first method is an augmented training approach where progressively more data from the new condition is mixed with the original data. The second method uses unsupervised adaptation to adapt the original model to the new conditions. Evaluation under changing environmental and noise conditions reveals the model produced from augmented training data to achieve higher detection accuracy across all conditions than the adapted model. However, the adapted model does not require label data from the new environment and in these situations is a more realistic solution.
@InProceedings{9287565,
author = {W. Vickers and B. Milner and A. Gorpincenko and R. Lee},
booktitle = {2020 28th European Signal Processing Conference (EUSIPCO)},
title = {Methods to Improve the Robustness of Right Whale Detection using CNNs in Changing Conditions},
year = {2020},
pages = {106-110},
abstract = {This paper is concerned with developing a method of detecting right whales from autonomous surface vehicles (ASVs) that is robust to changing operating conditions. A baseline convolutional neural network (CNN) is first trained using data taken from a single operating condition. Its detection accuracy is then found to degrade when applied to different operating conditions. Two methods are then investigated to restore performance using just a single model. The first method is an augmented training approach where progressively more data from the new condition is mixed with the original data. The second method uses unsupervised adaptation to adapt the original model to the new conditions. Evaluation under changing environmental and noise conditions reveals the model produced from augmented training data to achieve higher detection accuracy across all conditions than the adapted model. However, the adapted model does not require label data from the new environment and in these situations is a more realistic solution.},
keywords = {Training;Adaptation models;Whales;Training data;Signal processing;Data models;Surface treatment;cetacean detection;autonomous surface vehicles;passive acoustic monitoring;CNN;augmentation;adaptation},
doi = {10.23919/Eusipco47968.2020.9287565},
issn = {2076-1465},
month = {Aug},
url = {https://www.eurasip.org/proceedings/eusipco/eusipco2020/pdfs/0000106.pdf},
}
Downloads: 0
{"_id":"X5wkzzCeDDqpMMy7e","bibbaseid":"vickers-milner-gorpincenko-lee-methodstoimprovetherobustnessofrightwhaledetectionusingcnnsinchangingconditions-2020","authorIDs":[],"author_short":["Vickers, W.","Milner, B.","Gorpincenko, A.","Lee, R."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["W."],"propositions":[],"lastnames":["Vickers"],"suffixes":[]},{"firstnames":["B."],"propositions":[],"lastnames":["Milner"],"suffixes":[]},{"firstnames":["A."],"propositions":[],"lastnames":["Gorpincenko"],"suffixes":[]},{"firstnames":["R."],"propositions":[],"lastnames":["Lee"],"suffixes":[]}],"booktitle":"2020 28th European Signal Processing Conference (EUSIPCO)","title":"Methods to Improve the Robustness of Right Whale Detection using CNNs in Changing Conditions","year":"2020","pages":"106-110","abstract":"This paper is concerned with developing a method of detecting right whales from autonomous surface vehicles (ASVs) that is robust to changing operating conditions. A baseline convolutional neural network (CNN) is first trained using data taken from a single operating condition. Its detection accuracy is then found to degrade when applied to different operating conditions. Two methods are then investigated to restore performance using just a single model. The first method is an augmented training approach where progressively more data from the new condition is mixed with the original data. The second method uses unsupervised adaptation to adapt the original model to the new conditions. Evaluation under changing environmental and noise conditions reveals the model produced from augmented training data to achieve higher detection accuracy across all conditions than the adapted model. However, the adapted model does not require label data from the new environment and in these situations is a more realistic solution.","keywords":"Training;Adaptation models;Whales;Training data;Signal processing;Data models;Surface treatment;cetacean detection;autonomous surface vehicles;passive acoustic monitoring;CNN;augmentation;adaptation","doi":"10.23919/Eusipco47968.2020.9287565","issn":"2076-1465","month":"Aug","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2020/pdfs/0000106.pdf","bibtex":"@InProceedings{9287565,\n author = {W. Vickers and B. Milner and A. Gorpincenko and R. Lee},\n booktitle = {2020 28th European Signal Processing Conference (EUSIPCO)},\n title = {Methods to Improve the Robustness of Right Whale Detection using CNNs in Changing Conditions},\n year = {2020},\n pages = {106-110},\n abstract = {This paper is concerned with developing a method of detecting right whales from autonomous surface vehicles (ASVs) that is robust to changing operating conditions. A baseline convolutional neural network (CNN) is first trained using data taken from a single operating condition. Its detection accuracy is then found to degrade when applied to different operating conditions. Two methods are then investigated to restore performance using just a single model. The first method is an augmented training approach where progressively more data from the new condition is mixed with the original data. The second method uses unsupervised adaptation to adapt the original model to the new conditions. Evaluation under changing environmental and noise conditions reveals the model produced from augmented training data to achieve higher detection accuracy across all conditions than the adapted model. However, the adapted model does not require label data from the new environment and in these situations is a more realistic solution.},\n keywords = {Training;Adaptation models;Whales;Training data;Signal processing;Data models;Surface treatment;cetacean detection;autonomous surface vehicles;passive acoustic monitoring;CNN;augmentation;adaptation},\n doi = {10.23919/Eusipco47968.2020.9287565},\n issn = {2076-1465},\n month = {Aug},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2020/pdfs/0000106.pdf},\n}\n\n","author_short":["Vickers, W.","Milner, B.","Gorpincenko, A.","Lee, R."],"key":"9287565","id":"9287565","bibbaseid":"vickers-milner-gorpincenko-lee-methodstoimprovetherobustnessofrightwhaledetectionusingcnnsinchangingconditions-2020","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2020/pdfs/0000106.pdf"},"keyword":["Training;Adaptation models;Whales;Training data;Signal processing;Data models;Surface treatment;cetacean detection;autonomous surface vehicles;passive acoustic monitoring;CNN;augmentation;adaptation"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2020url.bib","creationDate":"2021-02-13T19:41:51.397Z","downloads":0,"keywords":["training;adaptation models;whales;training data;signal processing;data models;surface treatment;cetacean detection;autonomous surface vehicles;passive acoustic monitoring;cnn;augmentation;adaptation"],"search_terms":["methods","improve","robustness","right","whale","detection","using","cnns","changing","conditions","vickers","milner","gorpincenko","lee"],"title":"Methods to Improve the Robustness of Right Whale Detection using CNNs in Changing Conditions","year":2020,"dataSources":["wXzutN6o5hxayPKdC","NBHz6C7PWuqwYyaqa"]}