경첩 손실 함수 최소화를 통한 오디오 핑거프린트 이진화 = Audio Fingerprint Binarization by Minimizing Hinge-Loss Function

다국어 초록 (Multilingual Abstract)

This paper proposes a robust binary audio fingerprinting method by minimizing hinge-loss function. In the proposed method, the type of fingerprints is binary, which is conducive in reducing the size of fingerprint DB. In general, the binarization of features for fingerprinting deteriorates the performance of fingerprinting system, such as robustness and discriminability. Thus it is necessary to minimize such performance loss. Since the similarity between two audio clips is represented by a hinge-like function, we propose a method to derive a binary fingerprinting by minimizing a hinge-loss function. The derived hinge-loss function is minimized by using the minimal loss hashing. Experiments over thousands of songs demonstrate that the identification performance of binary fingerprinting can be improved by minimizing the proposed hinge loss function.

참고문헌 (Reference)

1 윤원중, "음악 정보검색 시스템을 위한 효율적인 특징 벡터 추출에 관한 연구" 한국음향학회 23 (23): 532-539, 2004

2 서진수, "압축 도메인 특징을 이용한 강인한 오디오 핑거프린팅" 한국음향학회 28 (28): 375-382, 2009

3 J. Herre., "Robust matching of audio signals using spectral latness features" 127-130, 2001

4 E. Zwicker, "Psychoacoustics: Facts and Models" Springer-Verlag 1999

5 D. Jang., "Pairwise Boosted Audio Fingerprint" 4 : 995-1004, 2009

6 M. Norouzi., "Minimal loss hashing for compact binary codes" 2011

7 M. Mohri., "Efficient and robust music identification with weighted finite-state transducers" 18 : 197-207, 2010

8 M. Casey., "Content-based music information retrieval : Current directions and future challenges" 96 : 668-696, 2008

9 Jin S. Seo., "Audio fingerprinting based on normalized spectral subband moments" 13 : 209-212, 2006

10 P. Indyk., "Approximate nearest neighbors: towards removing the curse of dimensionality" 1998

11 P. Cano., "A review of audio fingerprinting" 41 : 271-284, 2005

12 J. Haitsma., "A highly robust audio fingerprinting system" 2002