Deep neural networks for video classification in ecology

Simple item page
dc.contributor.advisorDurbach, Ian
dc.contributor.authorConway, Alexander
dc.date.accessioned2021-01-15T09:53:08Z
dc.date.available2021-01-15T09:53:08Z
dc.date.issued2020
dc.description.abstractAnalyzing large volumes of video data is a challenging and time-consuming task. Automating this process would very valuable, especially in ecological research where massive amounts of video can be used to unlock new avenues of ecological research into the behaviour of animals in their environments. Deep Neural Networks, particularly Deep Convolutional Neural Networks, are a powerful class of models for computer vision. When combined with Recurrent Neural Networks, Deep Convolutional models can be applied to video for frame level video classification. This research studies two datasets: penguins and seals. The purpose of the research is to compare the performance of image-only CNNs, which treat each frame of a video independently, against a combined CNN-RNN approach; and to assess whether incorporating the motion information in the temporal aspect of video improves the accuracy of classifications in these two datasets. Video and image-only models offer similar out-of-sample performance on the simpler seals dataset but the video model led to moderate performance improvements on the more complex penguin action recognition dataset.
dc.identifier.apacitationConway, A. (2020). <i>Deep neural networks for video classification in ecology</i>. (Master Thesis). University of Cape Town. Retrieved from http://hdl.handle.net/11427/32520en_ZA
dc.identifier.chicagocitationConway, Alexander. <i>"Deep neural networks for video classification in ecology."</i> Master Thesis., University of Cape Town, 2020. http://hdl.handle.net/11427/32520en_ZA
dc.identifier.citationConway, A. 2020. Deep neural networks for video classification in ecology. Master Thesis. University of Cape Town. http://hdl.handle.net/11427/32520en_ZA
dc.identifier.ris TY - Master Thesis AU - Conway, Alexander AB - Analyzing large volumes of video data is a challenging and time-consuming task. Automating this process would very valuable, especially in ecological research where massive amounts of video can be used to unlock new avenues of ecological research into the behaviour of animals in their environments. Deep Neural Networks, particularly Deep Convolutional Neural Networks, are a powerful class of models for computer vision. When combined with Recurrent Neural Networks, Deep Convolutional models can be applied to video for frame level video classification. This research studies two datasets: penguins and seals. The purpose of the research is to compare the performance of image-only CNNs, which treat each frame of a video independently, against a combined CNN-RNN approach; and to assess whether incorporating the motion information in the temporal aspect of video improves the accuracy of classifications in these two datasets. Video and image-only models offer similar out-of-sample performance on the simpler seals dataset but the video model led to moderate performance improvements on the more complex penguin action recognition dataset. DA - 2020 DB - OpenUCT DP - University of Cape Town LK - https://open.uct.ac.za PY - 2020 T1 - Deep neural networks for video classification in ecology TI - Deep neural networks for video classification in ecology UR - http://hdl.handle.net/11427/32520 ER - en_ZA
dc.identifier.urihttp://hdl.handle.net/11427/32520
dc.identifier.vancouvercitationConway A. Deep neural networks for video classification in ecology. [Master Thesis]. University of Cape Town, 2020 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/32520en_ZA
dc.language.isoeng
dc.publisherUniversity of Cape Town
dc.publisher.departmentDepartment of Statistical Sciences
dc.publisher.facultyFaculty of Science
dc.subject.otherDeep Neural Networks
dc.subject.otherComputer Vision
dc.titleDeep neural networks for video classification in ecology
dc.typeMaster Thesis
dc.type.qualificationlevelMasters
dc.type.qualificationnameMSc
uct.type.publicationResearch
uct.type.resourceMaster Thesis
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
thesis_sci_2020_conway_alexander.pdf
Size:
35.94 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Masters