DocumentCode :
2406290
Title :
Low Complexity H.264 Intra MB Coding
Author :
Kalva, Hari ; Kunzelmann, Philip ; Jillani, Rashad ; Pandya, Abhi
Author_Institution :
Dept. of Comput. Sci. & Eng., Florida Atlantic Univ., Boca Raton, FL
fYear :
2008
fDate :
9-13 Jan. 2008
Firstpage :
1
Lastpage :
2
Abstract :
The H.264 video coding standard is highly efficient and at the same time highly complex. The complexity of encoding is substantially higher than prior standards such as H.263 and makes H.264 video encoding on mobile devices expensive. The few existing encoding solutions on mobile devices restrict the features used there by sacrificing quality for complexity. Fast encoding algorithms are necessary to fully exploit the advanced compression features offered by H.264. In this paper we present a novel machine learning based approach to reducing the complexity of intra MB coding. The results show that machine learning has a great potential and can reduce the complexity substantially with negligible impact on quality. The results show that the proposed method reduces encoding time to about 1/3 of the reference implementation with a negligible loss in quality.
Keywords :
computational complexity; learning (artificial intelligence); video coding; H.264 video coding standard; fast encoding algorithms; intra coding; intra macro block coding; machine learning; macroblock encoding; mobile devices; video compression; Codecs; Computer science; Costs; Decision trees; Encoding; Machine learning; Machine learning algorithms; Standards development; Video coding; Video compression;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Consumer Electronics, 2008. ICCE 2008. Digest of Technical Papers. International Conference on
Conference_Location :
Las Vegas, NV
Print_ISBN :
978-1-4244-1458-1
Electronic_ISBN :
978-1-4244-1459-8
Type :
conf
DOI :
10.1109/ICCE.2008.4587996
Filename :
4587996
Link To Document :
بازگشت