Name
Title | ||
|---|---|---|
Efficient Orthogonal Matching Pursuit using sparse random projections for scene and video classification |
Abstract | ||
|---|---|---|
Sparse projection has been shown to be highly effective in several domains, including image denoising and scene / object classification. However, practical application to large scale problems such as video analysis requires efficient versions of sparse projection algorithms such as Orthogonal Matching Pursuit (OMP). In particular, random projection based locality sensitive hashing (LSH) has been proposed for OMP. In this paper, we propose a novel technique called Comparison Hadamard random projection (CHRP) for further improving the efficiency of LSH within OMP. CHRP combines two techniques:(1) The Fast Johnson-Lindenstrauss Transform (FJLT) which uses a randomized Hadamard transform and sparse projection matrix for LSH, and (2) Achlioptas' random projection that uses only addition and comparison operations. Our approach provides the robustness of FJLT while completely avoiding multiplications. We empirically validate CHRP's efficacy by performing a suite of experiments for image denoising, scene classification, and video categorization. Our experiments indicate that CHRP significantly speeds-up OMP with negligible loss in classification accuracy. |
Year | DOI | Venue |
|---|---|---|
2011 | 10.1109/ICCV.2011.6126512 | ICCV |
Keywords | Field | DocType |
speeds-up omp,object classification,comparison hadamard random projection,video classification,scene classification,classification accuracy,validate chrp,random projection,sparse projection matrix,sparse random projection,image denoising,sparse projection,efficient orthogonal matching pursuit,dictionaries,orthogonal matching pursuit,vectors,iterative methods,matching pursuit,artificial neural network,image classification,sparse matrices,locality sensitive hashing,artificial neural networks | Matching pursuit,Locality-sensitive hashing,Random projection,Pattern recognition,Computer science,Projection (linear algebra),Robustness (computer science),Artificial intelligence,Contextual image classification,Hadamard transform,Sparse matrix | Conference |
Volume | Issue | ISSN |
2011 | 1 | 1550-5499 |
Citations | PageRank | References |
12 | 0.71 | 23 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Shiv Naga Prasad Vitaladevuni | 1 | 272 | 18.18 |
| Premkumar Natarajan | 2 | 874 | 79.46 |
| Rohit Prasad | 3 | 465 | 39.06 |
| Prem Natarajan | 4 | 18 | 2.62 |