Correction to: Scientific Reports https://doi.org/10.1038/s41598-020-57549-y, published online 17 January 2020


In this Article, the legend for Table 1 is incorrect.


“Number of upcalls indicates the number of upcalls annotated by trained analysts. For deployments with two or more recorders, the number of upcalls indicates the total number of upcalls detected across all recorders. Shaded rows indicate data used to train neural networks. Non-shaded rows represent evaluation data. Negative examples for the Kaggle data represent the false detections flagged by the analysts as derived from non-right whale sources.”


should read:


“Data sources used to train and evaluate deep neural network performance. Number of upcalls indicates the number of upcalls annotated by trained analysts. For deployments with two or more recorders, it indicates the total number of upcalls detected across all recorders. Data from 28-Mar-09 to 31-Mar-09 were used to train neural networks. All others represent evaluation data. Negative examples for the Kaggle data represent the false detections flagged by the analysts as derived from non-right whale sources.”


In addition, the Acknowledgements section in this Article is incomplete.


“We are grateful to P. Dugan for running the BRP baseline detector; S. Kahl for sharing source code and advice on the methods; A. Rahaman, K. Hodge, B. Estabrook, D. Salisbury, M. Pitzrick, and C. Pelkie for helping with data analysis; F. Channell, C. Tessaglia-Hymes, and D. Jaskula for deploying and retrieving MARUs, and the DCLDE 2013 organizing committee. We thank the Bureau of Ocean Energy Management for the funding of MARU deployments, Excelerate Energy Inc. for the funding of Autobuoy deployment, and Michael J. Weise of the US Office of Naval Research for support (N000141712867).”


should read:


“We are grateful to P. Dugan for running the BRP baseline detector; S. Kahl for sharing source code and advice on the methods; A. Rahaman, K. Hodge, B. Estabrook, D. Salisbury, M. Pitzrick, and C. Pelkie for helping with data analysis; F. Channell, C. Tessaglia-Hymes, and D. Jaskula for deploying and retrieving MARUs, and the DCLDE 2013 organizing committee. We thank S. V. Parijs, G. Davis, C.W. Clark, L. Hatch, D. Wiley, and NOAA Fisheries for the DCLDE data and analyses. We thank the Bureau of Ocean Energy Management for the funding of MARU deployments, Excelerate Energy Inc. for the funding of Autobuoy deployment, and Michael J. Weise of the US Office of Naval Research for support (N000141712867).”