ResNet-based bio-acoustics presence detection technology of Hainan gibbon calls. (September 2022)
- Record Type:
- Journal Article
- Title:
- ResNet-based bio-acoustics presence detection technology of Hainan gibbon calls. (September 2022)
- Main Title:
- ResNet-based bio-acoustics presence detection technology of Hainan gibbon calls
- Authors:
- Ruan, Wenda
Wu, Keyi
Chen, Qingchun
Zhang, Chengyun - Abstract:
- Highlights: A ResNet based automated bio-acoustics presence detection frame-work was proposed. SpecAugment and label smoothing are effective in bio-acoustics presence detection task. An automatic post-processing method was proposed by employing the persistence characteristics. Abstract: As the most precious primate with only about 30 left, it is highly desirable to monitor the population of Hainan gibbon for the sake of effective protection. The traditional monitoring technique is facing the challenges of statistical hysteresis and a large number of labor costs, which makes it unsustainable. Deep learning based passive acoustic monitoring has shown its great potential in automatic detection of Hainan gibbon calls in Dufourq et al. (2021). Motivated by this progress, a bioacoustics presence detection network (BPDnet) was proposed for Hainan gibbon calls by tailoring acoustic signal pre-processing, feature extraction, label smoothing loss function, as well as data augmentation in the ResNet-based passive acoustic detection framework. In addition, a post-processing procedure was proposed to further improve the passive acoustic presence detection. Numerical results are presented to show that, the proposed BPDnet outperforms the baseline scheme of Dufourq et al. (2021) in terms of bio-acoustics presence detection accuracy, callback rate, F 1 -score by at least 12.9% when using the same 72-h test recording and test method without manual intervention. When manual post-processing isHighlights: A ResNet based automated bio-acoustics presence detection frame-work was proposed. SpecAugment and label smoothing are effective in bio-acoustics presence detection task. An automatic post-processing method was proposed by employing the persistence characteristics. Abstract: As the most precious primate with only about 30 left, it is highly desirable to monitor the population of Hainan gibbon for the sake of effective protection. The traditional monitoring technique is facing the challenges of statistical hysteresis and a large number of labor costs, which makes it unsustainable. Deep learning based passive acoustic monitoring has shown its great potential in automatic detection of Hainan gibbon calls in Dufourq et al. (2021). Motivated by this progress, a bioacoustics presence detection network (BPDnet) was proposed for Hainan gibbon calls by tailoring acoustic signal pre-processing, feature extraction, label smoothing loss function, as well as data augmentation in the ResNet-based passive acoustic detection framework. In addition, a post-processing procedure was proposed to further improve the passive acoustic presence detection. Numerical results are presented to show that, the proposed BPDnet outperforms the baseline scheme of Dufourq et al. (2021) in terms of bio-acoustics presence detection accuracy, callback rate, F 1 -score by at least 12.9% when using the same 72-h test recording and test method without manual intervention. When manual post-processing is allowed, the proposed BPDnet provides us an almost perfect Hainan gibbon call presence detection scheme. … (more)
- Is Part Of:
- Applied acoustics. Volume 198(2022)
- Journal:
- Applied acoustics
- Issue:
- Volume 198(2022)
- Issue Display:
- Volume 198, Issue 2022 (2022)
- Year:
- 2022
- Volume:
- 198
- Issue:
- 2022
- Issue Sort Value:
- 2022-0198-2022-0000
- Page Start:
- Page End:
- Publication Date:
- 2022-09
- Subjects:
- Bioacoustics presence detection -- Deep learning -- Hainan gibbon call
Acoustical engineering -- Periodicals
Periodicals
620.2 - Journal URLs:
- http://www.sciencedirect.com/science/journal/0003682X ↗
http://www.elsevier.com/journals ↗
http://www.elsevier.com/homepage/elecserv.htt ↗ - DOI:
- 10.1016/j.apacoust.2022.108939 ↗
- Languages:
- English
- ISSNs:
- 0003-682X
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 1571.400000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 23386.xml