Musical Feature Detection

1
Musical Feature Detection
(work in progress)

• Anton Eliens

2
Musical Feature Detection

• Introduction
• Architecture
• Extraction
• Query facilities
• Validation case study
• Open problems
• Conclusions

3
Introduction
gathering
query
description
extraction
similarity
4
Out there on the Web

• Aria Database title, category, voice part
• Midi Files on the Farm per genre
• Meta Searches AltaVista, Infoseek
• Lyrics search .

Keywords and categories
Btw. Why MIDI?
- Informix Musclefish Datablade - Meldex
www.nzdl.www
content
5
Architecture
6
Extraction - the anatomy of a midi file
7
detector song
to get the filename detector
lyrics
extracts lyrics detector melody
extracts
melody atom str name atom str text
atom str note midi song song file
lyrics melody file name lyrics
text melody note
Feature grammar
8
int melodyDetector(tree pt, list tks )
char buf1024 char _result void q
_query int idq 0 idq
query_eval(q,"Xmelody(X)") while ((_result
query_result(q,idq)) )
printf("note \s\n",_result)
putAtom(tks,"note",_result)
return SUCCESS
Melody detector
embedded logic
9
V1 newoid() midi_song.insert(oid(V0),oid(V
1)) V2 newoid() song_file.insert(oid(
V1),oid(V2)) file_name.insert(oid(V2),"ko
rtjakje") song_lyrics.insert(oid(V1),oid(V2
)) lyrics_text.insert(oid(V2),"e")
lyrics_text.insert(oid(V2),"per-")
lyrics_text.insert(oid(V2),"sonne")
lyrics_text.insert(oid(V2),"Moi")
lyrics_text.insert(oid(V2),"je")
lyrics_text.insert(oid(V2),"dis")
lyrics_text.insert(oid(V2),"que")
lyrics_text.insert(oid(V2),"les")
lyrics_text.insert(oid(V2),"bon-")
lyrics_text.insert(oid(V2),"bons")
lyrics_text.insert(oid(V2),"Val-")
lyrics_text.insert(oid(V2),"ent")
song_melody.insert(oid(V1),oid(V2))
melody_note.insert(oid(V2),"a-2")
melody_note.insert(oid(V2),"a-2")
melody_note.insert(oid(V2),"g-2")
melody_note.insert(oid(V2),"g-2")
melody_note.insert(oid(V2),"f-2")
melody_note.insert(oid(V2),"f-2")
melody_note.insert(oid(V2),"e-2")
melody_note.insert(oid(V2),"e-2")
melody_note.insert(oid(V2),"d-2")
melody_note.insert(oid(V2),"d-2")
melody_note.insert(oid(V2),"e-2")
melody_note.insert(oid(V2),"c-2")
Monet updates
Kortjakje.mid
10
extraction
11
Query
voice
12
Case study
Kortjakje
13
Representation
Song kortjakje
Composer Who cares.
Melody c c g g a a g g f f e e d d c
Score
14
Kortjakje variations Mozart
XII variations
15
melody transcription
Meldex
melody retrieval from tunes
16
Meldex
Hum that Tune
capture transcribe retrieval
Exact match
Approx match
www.nzdl.org
dynamic programming
17
No of notes rel. to size and alg.
Search times, fixed database size
18
Conclusions