Parallel Lasso for Large-Scale Video Concept Detection

Existing video concept detectors are generally built upon the kernel based machine learning techniques, e.g., support vector machines, regularized least squares, and logistic regression, just to name a few. However, in order to build robust detectors, the learning process suffers from the scalabilit...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on multimedia Vol. 14; no. 1; pp. 55 - 65
Main Authors: Geng, Bo, Li, Yangxi, Tao, Dacheng, Wang, Meng, Zha, Zheng-Jun, Xu, Chao
Format: Journal Article
Language:English
Published: New York, NY IEEE 01.02.2012
Institute of Electrical and Electronics Engineers
Subjects:
Algorithm design and analysis
Algorithmics. Computability. Computer arithmetics
Applied sciences
Artificial intelligence
Complexity theory
Computer science; control theory; systems
Computer systems and distributed systems. User interface
Educational institutions
Exact sciences and technology
Feature extraction
Incomplete cholosky factorization
Kernel
lasso
Learning and adaptive systems
Machine learning
Optimization
parallel learning
Pattern recognition. Digital image processing. Computational geometry
Software
Theoretical computing
video concept detection
Scalability
Video signal
Data transmission
Distributed computing
Modeling
Space complexity
Multidimensional analysis
Covariance
Cholesky method
Least squares method
Knowledge discovery
key frame
Learning algorithm
Cholesky factorization
Computer vision
video concept detection
Parallel algorithm
Statistical analysis
lasso
Interior point method
Computational complexity
Kernel method
parallel learning
Primal dual method
Parallel computation
Large scale
Incomplete cholosky factorization
Time complexity
Artificial intelligence
ISSN:1520-9210, 1941-0077
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Existing video concept detectors are generally built upon the kernel based machine learning techniques, e.g., support vector machines, regularized least squares, and logistic regression, just to name a few. However, in order to build robust detectors, the learning process suffers from the scalability issues including the high-dimensional multi-modality visual features and the large-scale keyframe examples. In this paper, we propose parallel lasso (Plasso) by introducing the parallel distributed computation to significantly improve the scalability of lasso (the l 1 regularized least squares). We apply the parallel incomplete Cholesky factorization to approximate the covariance statistics in the preprocess step, and the parallel primal-dual interior-point method with the Sherman-Morrison-Woodbury formula to optimize the model parameters. For a dataset with n samples in a d -dimensional space, compared with lasso, Plasso significantly reduces complexities from the original O ( d 3 ) for computational time and O ( d 2 ) for storage space to O ( h 2 d / m ) and O ( hd / m ) , respectively, if the system has m processors and the reduced dimension h is much smaller than the original dimension d . Furthermore, we develop the kernel extension of the proposed linear algorithm with the sample reweighting schema, and we can achieve similar time and space complexity improvements [time complexity from O ( n 3 ) to O ( h 2 n / m ) and the space complexity from O ( n 2 ) to O ( hn / m ), for a dataset with n training examples]. Experimental results on TRECVID video concept detection challenges suggest that the proposed method can obtain significant time and space savings for training effective detectors with limited communication overhead.
ISSN:1520-9210
1941-0077
DOI:10.1109/TMM.2011.2174781