Feature Vector Compression based on Least Error Quantization

被引:1
作者
Kawahara, Tomokazu [1 ]
Yamaguchi, Osamu [1 ]
机构
[1] Toshiba Co Ltd, Corp Res & Dev Ctr, Saiwai Ku, 1 Komukai Toshiba Cho, Kawasaki, Kanagawa 2128582, Japan
来源
PROCEEDINGS OF 29TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, (CVPRW 2016) | 2016年
关键词
D O I
10.1109/CVPRW.2016.18
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We propose a distinctive feature vector compression method based on least error quantization. This method can be applied to several biometrics methods using feature vectors, and allows us to significantly reduce the memory size of feature vectors without degrading the recognition performance. In this paper, we prove that minimizing quantization error between the compressed and original vectors is most effective to control the performance in face recognition. A conventional method uses non-uniform quantizer which minimizes the quantization error in terms of L-2-distance. However, face recognition methods often use metrics other than L-2-distance. Our method can calculate the quantized vectors in arbitrary metrics such as L-p-distance (0 < p <= infinity) and the quantized subspace basis. Furthermore, we also propose a fast algorithm calculating L-p-distances between two quantized vectors without decoding them. We evaluate the performance of our method on FERET, LFW and large face datasets with LBP (L-p-distance), Mutual Subspace Method and deep feature. The results show that the recognition rate using the quantized feature vectors is as accurate as that of the method using the original vectors even though the memory size of the vectors is reduced to 1/5 - 1/10. In particular, applying our method to the state-of-the-art feature, we are able to obtain the high performance feature whose size is very small.
引用
收藏
页码:84 / 92
页数:9
相关论文
共 30 条
  • [1] Aggarwal CC, 2001, LECT NOTES COMPUT SC, V1973, P420
  • [2] Face description with local binary patterns:: Application to face recognition
    Ahonen, Timo
    Hadid, Abdenour
    Pietikainen, Matti
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2006, 28 (12) : 2037 - 2041
  • [3] [Anonymous], 2005, IEEE COMP SOC C COMP
  • [4] [Anonymous], IEEE COMP SOC C COMP
  • [5] [Anonymous], INT C COMP VIS
  • [6] [Anonymous], 2007, Tech. rep
  • [7] Eigenfaces vs. Fisherfaces: Recognition using class specific linear projection
    Belhumeur, PN
    Hespanha, JP
    Kriegman, DJ
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1997, 19 (07) : 711 - 720
  • [8] Chandrasekhar Vijay, 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), P2504, DOI 10.1109/CVPRW.2009.5206733
  • [9] E. R. of Local Geometry for Large Scale Object Retrieval, 2009, IEEE COMP SOC C COMP, P9
  • [10] Fulkerson B, 2008, LECT NOTES COMPUT SC, V5302, P179, DOI 10.1007/978-3-540-88682-2_15