Western Indian Ocean blue whale dataset
The analysis of the large volumes of data resulting from continuous and long-term monitoring of blue whales unmistakably benefits from the automated detection of target signals. Automatic detection methods must be reliable and robust to gather statistically relevant elements and contribute to answer...
| Main Author: | |
|---|---|
| Format: | Dataset |
| Language: | unknown |
| Published: |
Zenodo
2020
|
| Subjects: | |
| Online Access: | https://dx.doi.org/10.5281/zenodo.3624145 https://zenodo.org/record/3624145 |
| Summary: | The analysis of the large volumes of data resulting from continuous and long-term monitoring of blue whales unmistakably benefits from the automated detection of target signals. Automatic detection methods must be reliable and robust to gather statistically relevant elements and contribute to answering blue whale ecological unknowns. Therefore, systematic performance analysis of detection and classification methods developed for the passive context is crucial. These datasets were annotated to support my Ph.D work Detection and classification in passive acoustic contexts: Application to blue whale low-frequency signals . They are based on recordings from Ocean Bottom Seismometers (RR41 and RR48) deployed during the RHUM-RUM experiment, in the western Indian Ocean. All RHUM-RUM recordings are freely available from the RESIF data center, under the code name YV (https://doi.org/10.15778/RESIF.YV2011). Find more information on RHUM-RUM on Researchgate. All blue whales calls within the provided acoustic recordings were manually annotated using the software Raven Pro 1.5. The analyst selected boxes describing the begin and end times and, minimum and maximum frequencies of each call (Dataset 1) or each unit (Dataset 2). Dataset 1 - Antarctic blue whale calls Contains: Sound file RR48_2013_D151.wav where Antarctic blue whale calls are recorded continuously for more than 21 hours (from 01:20 to 22:40); Annotation file Annotation_RR48_2013_D151.txt with the box information of 845 Antarctic blue whale calls (+ some other interfering events); and, Call time of arrivals estimated from events detected by the stochastic matched filter and corresponding whale location in the file Localisation_RR48_2013_D151.txt . This dataset supports research presented in: L. Bouffaut, R. Dréo, V. Labat, A. Boudraa and G. Barruol “Passive stochastic matched filter for antarctic blue whale call detection,” in J. Acoust. Soc. Am, vol. 144, no. 2, pp. 955-965 (2018). DOI: 10.1121/1.5050520 Additional information: - Detailed information on this dataset on my personal website - Stochastic matched filter package on Github DOI: 10.5281/zenodo.3613787 - Localization method research paper on Researchgate | Deep sea research paper Dataset 2 - Blue whale call units Contains: Sound files RR41_2013_D132.wav , RR41_2013_D135.wav , and RR48_2013_D138.wav , Corresponding annotation file Annotation.txt with box information of various blue whale call units from Antarctic blue whales, Madagascar pygmy blue whales and P-calls (+ some other interfering events). In total this dataset contains more than 4000 annotated units. This dataset supports research presented in: L. Bouffaut, S.Madhusudhana, V. Labat, A. Boudraa and H. Klinck, “Automated blue whale song transcription across variable acoustic contexts,” in IEEE OCEANS’19 Marseille, student poster competition, France, pp. 2-7 (2019) DOI: 10.1109/OCEANSE.2019.8867471 who received the First Prize and Norman Miller Award and was therefore also published in the IEEE OES Beacon Newsletter (September 2019 vol. 8, no. 3, pp. 44-49) Detailed information on this dataset on my personal website. |
|---|