Identifying illicit graphic in the online community using the neural network framework

Identifying illicit graphic in the online community using the neural network framework

In this paper two convolutional neural networks are estimated to classify whether an image contains a swastika or not. The images are gathered from the gaming platform Steam and by scraping a web search engine. The architecture of the networks is kept moderate and the difference between the models i...

Full description

Bibliographic Details
Main Author: Vega Ezpeleta, Emilio
Format: Others
Language:English
Published: Uppsala universitet, Statistiska institutionen 2017
Subjects:
Convolutional neural network
CNN
Neural networks
Image recognition
Statistical learning
Probability Theory and Statistics
Sannolikhetsteori och statistik
Online Access:http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-325810
id ndltd-UPSALLA1-oai-DiVA.org-uu-325810
record_format oai_dc
spelling ndltd-UPSALLA1-oai-DiVA.org-uu-3258102017-06-29T05:43:30ZIdentifying illicit graphic in the online community using the neural network frameworkengVega Ezpeleta, EmilioUppsala universitet, Statistiska institutionen2017Convolutional neural networkCNNNeural networksImage recognitionStatistical learningProbability Theory and StatisticsSannolikhetsteori och statistikIn this paper two convolutional neural networks are estimated to classify whether an image contains a swastika or not. The images are gathered from the gaming platform Steam and by scraping a web search engine. The architecture of the networks is kept moderate and the difference between the models is the final layer. The first model uses an average type operation while the second uses the conventional fully-connected layer at the end. The results show that the performance of the two models is similar and the test error is in the 6-9 % range. Student thesisinfo:eu-repo/semantics/bachelorThesistexthttp://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-325810application/pdfinfo:eu-repo/semantics/openAccess
collection NDLTD
language English
format Others
sources NDLTD
topic Convolutional neural network
CNN
Neural networks
Image recognition
Statistical learning
Probability Theory and Statistics
Sannolikhetsteori och statistik
spellingShingle Convolutional neural network
CNN
Neural networks
Image recognition
Statistical learning
Probability Theory and Statistics
Sannolikhetsteori och statistik
Vega Ezpeleta, Emilio
Identifying illicit graphic in the online community using the neural network framework
description In this paper two convolutional neural networks are estimated to classify whether an image contains a swastika or not. The images are gathered from the gaming platform Steam and by scraping a web search engine. The architecture of the networks is kept moderate and the difference between the models is the final layer. The first model uses an average type operation while the second uses the conventional fully-connected layer at the end. The results show that the performance of the two models is similar and the test error is in the 6-9 % range.
author Vega Ezpeleta, Emilio
author_facet Vega Ezpeleta, Emilio
author_sort Vega Ezpeleta, Emilio
title Identifying illicit graphic in the online community using the neural network framework
title_short Identifying illicit graphic in the online community using the neural network framework
title_full Identifying illicit graphic in the online community using the neural network framework
title_fullStr Identifying illicit graphic in the online community using the neural network framework
title_full_unstemmed Identifying illicit graphic in the online community using the neural network framework
title_sort identifying illicit graphic in the online community using the neural network framework
publisher Uppsala universitet, Statistiska institutionen
publishDate 2017
url http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-325810
work_keys_str_mv AT vegaezpeletaemilio identifyingillicitgraphicintheonlinecommunityusingtheneuralnetworkframework
_version_ 1718479790816624640

Similar Items