Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 The Representation, Indexing and Retrieval of Music Data at NTHU Arbee L.P. Chen National Tsing Hua University Taiwan, R.O.C.

Similar presentations


Presentation on theme: "1 The Representation, Indexing and Retrieval of Music Data at NTHU Arbee L.P. Chen National Tsing Hua University Taiwan, R.O.C."— Presentation transcript:

1 1 The Representation, Indexing and Retrieval of Music Data at NTHU Arbee L.P. Chen National Tsing Hua University Taiwan, R.O.C. http://www.cs.nthu.edu.tw/~alpchen

2 2 Outline Content-based media data retrieval Music data retrieval Features of music data Feature indexing and matching Prototypes Reference

3 3 Content-based Media Data Retrieval Representation of media contents features Feature extraction from media data Feature indexing Query interface

4 4 Content-based Media Data Retrieval Matching query features against the feature index approximate/partial matching similarity measure precision: how many of the answers are in fact correct recall: how many of the correct answers are in fact retrieved relevance feedback

5 5 Music Data Retrieval: System Architecture

6 6 Features of Music Data

7 7 Static music information The intrinsic music characteristics of music objects Key, beat, and tempo E.g., the Beethoven Symphony No. 5, Op. 67, C minor, 4/4, Allegro con brio Acoustical features Loudness, pitch, duration, bandwidth and brightness Can be computed and represented as numerical values

8 8 Features of Music Data Thematic features Themes, melodies, rhythms, and chords Can be derived from the staff information of a music object Melody The melody of a song is the sequence of the pitches of all notes in the songs E.g., the melody of the theme of the Beethoven ’ s Symphony No.5 is “ sol – sol – sol – mi – fa – fa – fa - re ”

9 9 Features of Music Data Rhythm The rhythm of a song is the sequence of the durations of all notes in the songs E.g., the rhythm of the theme of the Beethoven ’ s Symphony No.5 is “ 1/2-1/2-1/2-2-1/2-1/2-1/2-4 ” Chord A chord consists of three (root, third, and fifth) or more notes which sound together in harmony

10 10 Features of Music Data Coding scheme: a music object  a sequence of music segments music segment = (segment type, segment duration, segment pitch) four segment types: ┌┐ (type A), └┘ (type B), ┌┘ (type C), and └┐ (type D)

11 11 Features of Music Data For example, the sequence of music segments: (B,3,-3) (A,1,+1) (D,3,-3) (B,1,-2) (C,1,+2) (C,1,+2) (C,1,+1)

12 12 music segment = (type, duration, pitch)

13 13 Features of Music Data Repeating Pattern A sequence of notes appearing more than once in the music object Efficient content-based retrieval Semantics-rich representation Extracting repeating patterns Tree-based approach Matrix-based approach

14 14 Features of Music Data Experiment 1

15 15 Features of Music Data Dissimilarity of melody strings

16 16 Features of Music Data Dissimilarity of repeating patterns

17 17 Features of Music Data Experiment 2

18 18 Features of Music Data Validity of classes

19 19 Finding Repeating Patterns: Tree-based Approach Construct an RP-tree for RP ’ s with lengths 2 n, n  0, 1,... S = “ ABCDEFGHABCDEFGHIJABC ”

20 20 Finding Repeating Patterns: Tree-based Approach Length 1 {A, 3, (1, 9, 19)} {B, 3, (2, 10, 20)} {C, 3, (3, 11, 21)} {D, 2, (4, 12)} {E, 2, (5, 13)} {F, 2, (6, 14)} {G, 2, (7, 15)} {H, 2, (8, 16)}

21 21 Finding Repeating Patterns: Tree-based Approach Length 2 {AB, 3, (1, 9, 19)} = {A, 3, (1, 9, 19)}  0 {B, 3, (2, 10, 20)} {BC, 3, (2, 10, 20)} = {B, 3, (2, 10, 20)}  0 {C, 3, (3, 11, 21)} {CD, 2, (3, 11)} = {C, 3, (3, 11, 21)}  0 {D, 2, (4, 12)} …

22 22 Finding Repeating Patterns: Tree-based Approach Length 4 {ABCD, 2, (1, 9)} = {AB, 3, (1, 9, 19)}  0 {CD, 2, (3, 11)} {BCDE, 2, (2, 10)} = {BC, 2, (2, 10, 20)}  0 {DE, 2, (4, 12)} … Length 8 {ABCDEFGH, 2, (1, 9)} = {ABCD, 2, (1, 9)}  0 {EFGH, 2, (5, 13)}

23 23 Finding Repeating Patterns: Tree-based Approach

24 24 Finding Repeating Patterns: Tree-based Approach Prune trivial patterns of length 2 n, n = 0, 1, … Let X be an RP of S, Y a substring of X, and Z a substring of Y If freq(X) = freq(Z), Y is trivial

25 25 Finding Repeating Patterns: Tree-based Approach Length 1 {ABCDEFGH, 2, (1, 9)} {ABCD, 2, (1, 9)} {BCDE, 2, (2, 10)} {CDEF, 2, (3, 11)} {DEFG, 2, (4, 12)} {EFGH, 2, (5, 13)} {AB, 3, (1, 9, 19)} {BC, 3, (2, 10, 20)} {CD, 2, (3, 11)} {DE, 2, (4, 12)} {EF, 2, (5, 13)} {FG, 2, (6, 14)} {GH, 2, (7, 15)}

26 26 Finding Repeating Patterns: Tree-based Approach Length 2 {ABCDEFGH, 2, (1, 9)} {ABCD, 2, (1, 9)} {BCDE, 2, (2, 10)} {CDEF, 2, (3, 11)} {DEFG, 2, (4, 12)} {EFGH, 2, (5, 13)} {AB, 3, (1, 9, 19)} {BC, 3, (2, 10, 20)}

27 27 Finding Repeating Patterns: Tree-based Approach Length 4 {ABCDEFGH, 2, (1, 9)} {AB, 3, (1, 9, 19)} {BC, 3, (2, 10, 20)}

28 28 Finding Repeating Patterns: Tree-based Approach Generate all patterns of lengths  2 n, n  0, 1,... {ABCDEFGH, 2, (1, 9)} {AB, 3, (1, 9, 19)} {BC, 3, (2, 10, 20)} {ABC, 3, (1, 9, 19)} order-1 string-join AB  1 BC = ABC

29 29 Finding Repeating Patterns: Tree-based Approach Prune all trivial patterns {ABCDEFGH, 2, (1, 9)} {ABC, 3, (1, 9, 19)}

30 30 Feature Indexing and Matching 1D-List PAT-Tree L-Tree Augmented Suffix Tree Grid-Twin Suffix Tree

31 31 Feature Indexing and Matching: 1D-List There are two music objects M1 and M2 M1: ” sol-mi-mi-fa-re-re-do-re-mi-fa-sol-sol- sol ” M2: ” do-mi-sol-sol-re-mi-fa-fa-do-re-re-mi ” The melody string of the music query Q: ” do-re-mi ” Problem: to find whether M1 and M2 contain the melody string Q

32 32

33 33

34 34 Feature Indexing and Matching: PAT-Tree Example, songs in chord strings Song1 : Am F2 Dm Am Song2 : C C F C Song3 : G E1 C D Song4 : E1 G Am Bm

35 35 Feature Indexing and Matching: PAT-Tree

36 36 Prototype 1

37 37 Prototype 1

38 38 Prototype 2

39 39 Prototype 2

40 40 Prototype 2

41 41 References (http://db.nthu.edu.tw) Chen, A.L.P., M. Chang, J. Chen, J.L. Hsu, C.H. Hsu, and S.Y.S. Hua, “ Query by Music Segments:An Efficient Approach for Song Retrieval, ” in Proc. of IEEE Intl. Conference on Multimedia and Expo, 2000. Chen, J.C.C. and A.L.P. Chen, “ Query by Rhythm:An Approach for Song Retrieval in Music Database, ” in Proc. of IEEE Intl. Workshop on Research Issues in Data Engineering, 1998. Chou, T.C., A.L.P. Chen, and C.C. Liu, “ Music Databases: Indexing Techniques and Implementation, ” in Proc. of IEEE Intl. Workshop on Multimedia Data Base Management System, 1996. Hsu, J.L., C.C. Liu, and A.L.P. Chen, “ Efficient Repeating Pattern Finding in Music Databases, ” in Proc. of ACM Intl. Conference on Information and Knowledge Management, 1998.

42 42 References (http://db.nthu.edu.tw) Lee, W and A.L.P. Chen, “ Efficient Multi-Feature Index Structure for Music Data Retrieval, ” in Proc. of SPIE Conference on Storage and Retrieval for Image and Video Databases, 2000. Liu, C.C., J.L. Hsu, and A.L.P. Chen, “ An Approximate String Matching Algorithm for Content-Based Music Data Retrieval, ” in Proc. of IEEE Intl.Conference on Multimedia Computing and Systems, 1999. Liu, C.C., J.L. Hsu, and A.L.P. Chen, “ Efficient Theme and Non- Trivial Repeating Pattern Discovering in Music Databases, ” in Proc. of IEEE Intl. Conference on Data Engineering, 1999.


Download ppt "1 The Representation, Indexing and Retrieval of Music Data at NTHU Arbee L.P. Chen National Tsing Hua University Taiwan, R.O.C."

Similar presentations


Ads by Google