North Atlantic right whale localization and recognition using very deep and leaky Neural Network

dc.contributor.authorKabani, A.
dc.contributor.authorEl-Sakka, M. R.
dc.coverage.issue2cs
dc.coverage.volume5cs
dc.date.accessioned2017-02-03T11:15:56Z
dc.date.available2017-02-03T11:15:56Z
dc.date.issued2016cs
dc.description.abstractWe describe a deep learning model that can be used to recognize individual right whales in aerial images. We developed our model using a data set provided by the National Oceanic and Atmospheric Administration. The main challenge we faced when working on this data set is that the size of the training set is very small (4,544 images) with some classes having only 1 image. While this data set is by far the largest of its kind, it is very di cult to train a deep neural network with such a small data set. However, we were able to overcome this challenge by dividing this problem into smaller tasks and by reducing the viewpoint variance in the data set. First, we localize the body and the head of the whale using deep learning. Then, we align the whale and normalize it with respect to rotation. Finally, a network is used to recognize the whale by analyzing its callosities. The top-1 accuracy of the model is 69.7% and the top-5 accuracy is 85%. The solution we describe in this paper was ranked 5th (out of 364 teams) in a challenge to solve this problem.en
dc.formattextcs
dc.format.extent155-170cs
dc.format.mimetypeapplication/pdfen
dc.identifier.citationMathematics for Applications. 2016 vol. 5, č. 2, s. 155-170. ISSN 1805-3629cs
dc.identifier.doi10.13164/ma.2016.11en
dc.identifier.issn1805-3629
dc.identifier.urihttp://hdl.handle.net/11012/63787
dc.language.isoencs
dc.publisherVysoké učení technické v Brně, Fakulta strojního inženýrství, Ústav matematikycs
dc.relation.ispartofMathematics for Applicationsen
dc.relation.urihttp://ma.fme.vutbr.cz/archiv/5_2/ma_5_2_kabani_el_sakka_final.pdfcs
dc.rights© Vysoké učení technické v Brně, Fakulta strojního inženýrství, Ústav matematikycs
dc.rights.accessopenAccessen
dc.subjectwhale localizationen
dc.subjectwhale detectionen
dc.subjectwhale recognitionen
dc.subjectdeep learningen
dc.subjectconvolutional neural networken
dc.subjectlocalizationen
dc.subjectdetectionen
dc.subjectrecognitionen
dc.subjectimage classi cationen
dc.titleNorth Atlantic right whale localization and recognition using very deep and leaky Neural Networken
dc.type.driverotheren
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen
eprints.affiliatedInstitution.departmentÚstav matematikycs
eprints.affiliatedInstitution.facultyFakulta strojního inženýrstvícs
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
ma_5_2_kabani_el_sakka_final.pdf
Size:
17.25 MB
Format:
Adobe Portable Document Format
Description:
Collections