본문바로가기

자료 카테고리

전체 1
도서자료 0
학위논문 1
연속간행물·학술기사 0
멀티미디어 0
동영상 0
국회자료 0
특화자료 0

도서 앰블럼

전체 (0)
일반도서 (0)
E-BOOK (0)
고서 (0)
세미나자료 (0)
웹자료 (0)
전체 (1)
학위논문 (1)
전체 (0)
국내기사 (0)
국외기사 (0)
학술지·잡지 (0)
신문 (0)
전자저널 (0)
전체 (0)
오디오자료 (0)
전자매체 (0)
마이크로폼자료 (0)
지도/기타자료 (0)
전체 (0)
동영상자료 (0)
전체 (0)
외국법률번역DB (0)
국회회의록 (0)
국회의안정보 (0)
전체 (0)
표·그림DB (0)
지식공유 (0)

도서 앰블럼

전체 1
국내공공정책정보
국외공공정책정보
국회자료
전체 ()
정부기관 ()
지방자치단체 ()
공공기관 ()
싱크탱크 ()
국제기구 ()
전체 ()
정부기관 ()
의회기관 ()
싱크탱크 ()
국제기구 ()
전체 ()
국회의원정책자료 ()
입법기관자료 ()

검색결과

검색결과 (전체 1건)

검색결과제한

열기
논문명/저자명
A Highly robust audio fingerprinting scheme in real environments = 실제 환경에 강인한 오디오 핑거프린팅 기법 / 박만수 인기도
발행사항
대전 : 한국정보통신대학교 대학원, 2006.8
청구기호
TD 621.3828 ㅂ167h
형태사항
xii, 88 p. ; 26 cm
자료실
전자자료
제어번호
KDMT1200686759
주기사항
학위논문(박사) -- 한국정보통신대학교 대학원, 공학, 2006.8
원문
미리보기

목차보기더보기

Title page

Abstract

Contents

1. Introduction 18

1.1 Research Motivation 18

1.2 Research Approach 21

1.3 Organization 24

2. Stochastic Modeling Scheme 25

2.1 Spectral Envelope Histogram 25

2.2 Pitch Histogram 28

2.3 Dynamic Matching Method 32

2.4 Distance Metric 34

3. Audio Hashing Scheme 37

3.1 Overview 37

3.2 Frequency-Temporal Filtering 40

3.3 Bit Allocation 47

3.4 Two-Step Audio Hashing 50

3.5 Database Structure 52

3.5.1 Database Layout 52

3.5.2 Database Lookup 54

4. Experiments 58

4.1 Audio Data 58

4.2 Performance Evaluation of Stochastic Modeling Scheme 62

4.2.1 Distance Metric 62

4.2.2 Spectral Envelope Histogram vs. Pitch Histogram 64

4.2.3 Static Matching Method vs. Dynamic Matching Method 68

4.3 Performance Evaluation of Audio Hashing Scheme 70

4.3.1 Frequency-Temporal Filtering 70

4.3.2 Bit Allocation 78

4.4 Performance Evaluation in Real Environments 80

4.4.1 Time Stretch Modifications 80

4.4.2 Real-Noise Environments 83

5. Conclusions and Further Studies 88

국문요약 90

Reference 95

Acknowledgement 100

Table 4.1. Performance comparison of various distance metrics for spectral envelope histogram. 63

Table 4.2. Performance comparison of various cut off frequencies for pitch detection in pitch histogram. 65

Table 4.3. Performance evaluation of the static and dynamic matching methods in the stochastic modeling scheme based on pitch histogram. 69

Table 4.4. Performance evaluation of alternatives to frequency filtering when HT1 is used as a temporal filter. 72

Table 4.5. Performance evaluation of alternatives to temporal filtering when HT1 is used as a frequency filter. 73

Table 4.6. Performance comparison according to the frequency-temporal filtering combinations when the lookup candidates are not expanded (HD=0). 75

Table 4.7. Performance comparison according to the frequency-temporal filtering combinations when the lookup candidates are expanded (HD≤1). 76

Table 4.8. Performance evaluation of the two-step audio hashing scheme by frequency-temporal filtering. 77

Table 4.9. Performance evaluation according to bit allocation methods. 79

Table 4.10. Performance evaluation of tempo-change resilience of audio fingerprints. 82

Figure 2.1. Stochastic modeling scheme based on spectral envelope histogram. 27

Figure 2.2. Stochastic modeling scheme based on pitch histogram. 31

Figure 2.3. Dynamic matching process. 33

Figure 3.1. Overview of Philips audio fingerprinting scheme. 39

Figure 3.2. Filter response of various frequency filters. 42

Figure 3.3. Signal corruption by noise. 43

Figure 3.4. Filter response of various temporal filters. 46

Figure 3.5. Audio hashing scheme by optimal quantizer. 49

Figure 3.6. Two-step audio hashing scheme. 51

Figure 3.7. Database layout of an audio hashing scheme. 53

Figure 3.8. Overview of database lookup for an audio hashing scheme. 56

Figure 3.9. Database lookup process for the two-step audio hashing scheme. 57

Figure 4.1. Power spectrums of audio signals over various loudspeakers. 61

Figure 4.2. Performance comparison between spectral envelope histogram and pitch histogram corresponding to the number of histogram bins. 67

Figure 4.3. Performance comparison by a type of audio fingerprints for different SNR at various recording locations. 85

Figure 4.4. Performances of the frequency-temporal filtering combination for different SNR at various recording locations. 86

Figure 4.5. Performance evaluation of the two-step audio hashing scheme by bit allocation methods. 87

권호기사보기

권호기사 목록 테이블로 기사명, 저자명, 페이지, 원문, 기사목차 순으로 되어있습니다.
기사명 저자명 페이지 원문 기사목차
연속간행물 팝업 열기 연속간행물 팝업 열기