Structure-Based Learning in Sampling, Representation and Analysis for Multimedia Big Data

Author

Hongkai Xiong

Author_Institution

Dept. of Electron. Eng., Shanghai Jiao Tong Univ., Shanghai, China

fYear

2015

fDate

20-22 April 2015

Firstpage

24

Lastpage

27

Abstract

This paper presents disruptive insights and techniques on structure-based learning for multimedia big data. Along this viewpoint, significant technical challenges for multimedia big data are investigated, including sampling and reconstruction, representation, and analysis. For multi-dimensional compressive sampling, the union of data-driven subspace is addressed via subspace learning with structured sparsity. To enrich the correlated reconstruction, spatio-temporal regularity is presented within various multimedia data. Inspired by this insight, multi-scale dictionary learning is proposed to leverage spatio-temporal structures for sparse representation and make learning-based structured prediction and analysis.

Keywords

Big Data; data analysis; data structures; learning (artificial intelligence); multimedia computing; sampling methods; correlated reconstruction; data reconstruction; data-driven subspace; learning-based structured prediction; multi-scale dictionary learning; multidimensional compressive sampling; multimedia big data analysis; multimedia big data representation; multimedia big data sampling; multimedia data; sparse representation; spatio-temporal regularity; spatio-temporal structures; structure-based learning; structured sparsity; subspace learning; Big data; Dictionaries; Multimedia communication; Streaming media; Video coding; Videos; Visualization; Human action recognition; classifier; depth motion map; histogram of oriented gradients;

fLanguage

English

Publisher

ieee

Conference_Titel

Multimedia Big Data (BigMM), 2015 IEEE International Conference on

Conference_Location

Beijing

Print_ISBN

978-1-4799-8687-3

Type

conf

DOI

10.1109/BigMM.2015.83

Filename

7153771