Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
THE IMPACT OF USING GENERATED DATA IN LEARNING OBJECT DETECTION
Umeå University, Faculty of Science and Technology, Department of Computing Science.
2019 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

‘The development of autonomous cars is in full progress. For the autonomous cars to be able to detect where other cars are, many machine learning models are utilised. One problem in the fi€eld of object detection, is that a human has to tell where objects are in images, for a machine to be able to learn to detect objects in images it has never seen before. To tackle this problem synthetic images can be created, where the ground truth of where objects are in images is known, without having to use human knowledge. Th‘is thesis study the approach of using a translation model to translate images to look more like real photographs. Several object detection models are then used to evaluate if training on the translated images increases the generalisation to the real world photographs. In some cases the results show that the translated images help to increase the performance on real world photographs when detecting cars in images of street scenes.

Place, publisher, year, edition, pages
2019. , p. 24
Series
UMNAD ; 1193
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:umu:diva-163546OAI: oai:DiVA.org:umu-163546DiVA, id: diva2:1354434
Educational program
Bachelor of Science Programme in Computing Science
Supervisors
Examiners
Available from: 2019-09-25 Created: 2019-09-25 Last updated: 2019-09-25Bibliographically approved

Open Access in DiVA

fulltext(12484 kB)6 downloads
File information
File name FULLTEXT01.pdfFile size 12484 kBChecksum SHA-512
14c7ee741666d0ff214a04ef8c0125a532efa27924bd720b487cd216edb7afba66932707d59f38fad41c79750b24bfca6f630124cc058871846b2be2838c9db4
Type fulltextMimetype application/pdf

By organisation
Department of Computing Science
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 6 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 26 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf