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
Video content annotation automation using machine learning techniques
KTH, School of Computer Science and Communication (CSC).
2014 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Automatisk annotering av videomaterial med hjalp av maskininlarningstekniker (Swedish)
Abstract [en]

Annotations describing semantic content within video material is essential for ecient search of such content, e.g. allowing for search engine queries to return only relevant segments of video clips within large content management systems. However, manual annotation of video material is a dull and time-consuming task, eectively lowering thequality and quantity of such annotations. In this rapport a system to automate most of the process is suggested. The system learns from video material with user-provided annotationsto infer annotations for new material automatically, without requiring any system changes between dierent user-created labeling schemes. The prototype of such a system presentedin this rapport, evaluated on a few concepts, is showing promising results for concepts with high inuence on the scene environments.

Abstract [sv]

Annotering av semantiskt innehall av videomaterial ar kritiskt for eektiv sokning inomsadant material, vilket i sin tur mojliggor t.ex. att forfragningar till sokmotorer kanreturerna endast relevanta segment av videoklipp inom stora videohanteringssystem.Manuell annotering ar dock en trakig och tidsodande uppgift, vilket medfor lag kvalite ochliten mangd av sadana annoteringar. I denna uppsats foreslas ett system for attautomatisera det mesta av den processen. Systemet lar sig fran manuellt annoteratvideomaterial att inferera annoteringar for nytt material automatiskt, utan att kravaandringar pa systemet mellan olika anvandarskapta koncept att annotera. Prototypen somar presenterad i denna uppsats och utvarderad pa ett antal koncept visar lovande resultatfor koncept som har hogt inytande pa scenmiljoerna.

Place, publisher, year, edition, pages
2014.
National Category
Computer and Information Science Computer Science
Identifiers
URN: urn:nbn:se:kth:diva-155751OAI: oai:DiVA.org:kth-155751DiVA: diva2:762678
Educational program
Master of Science - Computer Science
Examiners
Available from: 2014-11-20 Created: 2014-11-12 Last updated: 2014-11-20Bibliographically approved

Open Access in DiVA

fulltext(5376 kB)186 downloads
File information
File name FULLTEXT01.pdfFile size 5376 kBChecksum SHA-512
83d17bc1ba4c5e96c584814ca61950de48822b002ef4b261eab009adcc352fb91bf2f2148eb4de91068749d34722ce1ae9b1c0e37846a4856bf4575aec0a8337
Type fulltextMimetype application/pdf

By organisation
School of Computer Science and Communication (CSC)
Computer and Information ScienceComputer Science

Search outside of DiVA

GoogleGoogle Scholar
Total: 186 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: 868 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