Presentation | 2014-03-13 Neighbor-to-Neighbor Search for Fast Image Classification Nakamasa INOUE, Koichi SHINODA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We propose Neighbor-to-Neighbor (NTN) search for fast image classification. The NTN search reduces computational cost of vector quantization (VQ) and Gaussian mixture models (GMMs) in image classification frame-works such as bag-of-visual-words and Fisher vector. The NTN search finds similar input vectors from a neighbor vector to a neighbor vector to skip some calculations based on the similarity of the input vectors. For example, in dense SIFT, the NTN search seeks similar descriptors from an adjacent descriptor to the other adjacent descriptors. We evaluated our method on the PASCAL VOC 2007 classification challenge task. The NTN search for VQ reduced the computational cost by 77.4%, and the NTN search for GMM reduced it by 89.3%, without any significant degradation in classification performance. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Image Classification / Gaussian Mixture Models / Bag-of-visual-words |
Paper # | PRMU2013-184 |
Date of Issue |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2014/3/6(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Pattern Recognition and Media Understanding (PRMU) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Neighbor-to-Neighbor Search for Fast Image Classification |
Sub Title (in English) | |
Keyword(1) | Image Classification |
Keyword(2) | Gaussian Mixture Models |
Keyword(3) | Bag-of-visual-words |
1st Author's Name | Nakamasa INOUE |
1st Author's Affiliation | Department of Computer Science, Tokyo Institute of Technology() |
2nd Author's Name | Koichi SHINODA |
2nd Author's Affiliation | Department of Computer Science, Tokyo Institute of Technology |
Date | 2014-03-13 |
Paper # | PRMU2013-184 |
Volume (vol) | vol.113 |
Number (no) | 493 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |