Table 13 The best retrieval performances obtained by ten consecutive SLP classifiers
From: An evolutionary feature synthesis approach for content-based audio retrieval
FS | Basic database (6 classes) | Extended database (12 classes) | ||
|---|---|---|---|---|
ANMRR (run no.) | AP (%) (run no.) | ANMRR (run no.) | AP (%) (run no.) | |
Segment | ||||
STAT | 0.206 (run 3) | 78.6 (run 3) | 0.579 (run 2) | 40.7 (run 2) |
MFCC | 0.109 (run 3) | 89.7 (run 3) | 0.479 (run 2) | 51.2 (run 2) |
Δ-MFCC | 0.295 (run 2) | 69.8 (run 2) | 0.534 (run 3) | 45.2 (run 3) |
ΔΔ-MFCC | 0.269 (run 10) | 72.4 (run 10) | 0.579 (run 4) | 41.3 (run 4) |
LPC | 0.483 (run 2) | 50.9 (run 2) | 0.667 (run 4) | 32.3 (run 4) |
LPCC | 0.550 (run 2) | 44.2 (run 2) | 0.714 (run 1) | 27.4 (run 1) |
S_AUDIO | 0.202 (run 5) | 79.4 (run 5) | 0.541 (run 2) | 44.7 (run 2) |
Key-frame | ||||
MFCC + deltas | 0.279 (run 1) | 70.2 (run 1) | 0.414 (run 1) | 56.5 (run 1) |
LPC + LPCC | 0.669 (run 1) | 32.1 (run 1) | 0.794 (run 1) | 19.7 (run 1) |
K_AUDIO | 0.277 (run 1) | 69.6 (run 1) | 0.504 (run 1) | 47.1 (run 1) |