mdh.sePublications
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
OBJECT RECOGNITION THROUGH CONVOLUTIONAL LEARNING FOR FPGA
Mälardalen University, School of Innovation, Design and Engineering.
2017 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

In later years the interest for deep networks and convolutional networks in regards to object recognition has spiked. There are however not many that focuses on the hardware in these subjects. This thesis was done in collaboration with Unibap to explore the feasibility of implementing these on a FPGA to speed up object recognition. A custom database is created to investigate if a smaller database could be utilized with good results. This database alongside the MNIST database are tested in multiple configurations to find a suitable solution with good enough accuracy. This thesis focuses on getting an accuracy which could be applicable in industries of today and is therefore not as driven by accuracy as many other works. Furthermore a FPGA implementation that is versatile and flexible enough to utilize regardless of network configuration is built and simulated. To achieve this research was done on existing AI and the focus landed on convolutional neural networks. The different configurations are all presented in regards to time, resource utilization and accuracy. The FPGA implementation in this work is only simulated and this leaves the desire and need to syntethize it on an actual FPGA.

Place, publisher, year, edition, pages
2017. , 31 p.
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:mdh:diva-36039OAI: oai:DiVA.org:mdh-36039DiVA: diva2:1119717
External cooperation
Unibap AB
Supervisors
Available from: 2017-08-29 Created: 2017-07-04 Last updated: 2017-08-29Bibliographically approved

Open Access in DiVA

fulltext(671 kB)24 downloads
File information
File name FULLTEXT01.pdfFile size 671 kBChecksum SHA-512
47b01bd8ed427d77f944a374787a94cc113122ee75b78dfd96b7764f07ae734beb62766c5b908d51e2cbdc6e8041a8c523ed00a19943b62eca4ab6161cd2e1d7
Type fulltextMimetype application/pdf

By organisation
School of Innovation, Design and Engineering
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 24 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

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