KNOWLEDGE TRANSFER FROM WEAKLY LABELED AUDIO USING CONVOLUTIONAL NEURAL NETWORK FOR SOUND EVENTS AND SCENES. pdf

Authors: Anurag Kumar, Maksim Khadkevich, Christian Fügen


Area Under ROC Curves (AUC)

average Precision

Figure compared AP of \(\mathcal{N}_S\) and \(\mathcal{N}_S^{slat}\). The Index of sound events 1 to 50 is shown below. The first column is Index in the above figure, second is the event id (magenta) as used in Audioset dataset and third (blue) is sound event name

1    /m/0d8_n   Pizzicato
2    /m/09ct_   Helicopter
3    /m/0912c9   Vehicle horn- car horn- honking
4    /m/0283d   Drum and bass
5    /t/dd00134   Car passing by
6    /m/07gql   Trumpet
7    /m/07rc7d9   Bow-wow
8    /m/03qtq   Hi-hat
9    /m/028sqc   Music of Asia
10    /m/01yrx   Cat
11    /m/02x8m   Funk
12    /m/05r6t   Punk rock
13    /t/dd00034   Tender music
14    /m/0dwtp   Glockenspiel
15    /m/026z9   Disco
16    /m/0g293   Music of Latin America
17    /m/06bz3   Radio
18    /m/032s66   Gunshot- gunfire
19    /t/dd00067   Heavy engine (low frequency)
20    /m/016cjb   Gospel music
21    /m/02cjck   Theme music
22    /m/053hz1   Cheering
23    /m/06h7j   Run
24    /m/08cyft   Electronic dance music
25    /m/07lnk   Trance music
26    /m/06bxc   Rapping
27    /m/03t3fj   Rimshot
28    /m/0ch8v   Livestock- farm animals- working animals
29    /m/07pggtn   Chirp- tweet
30    /m/01s0ps   Electric piano
31    /m/01qbl   Cymbal
32    /m/0155w   Blues
33    /m/02rr_   Effects unit
34    /m/0l14j_   Flute
35    /m/01c194   Mantra
36    /m/03_d0   Jazz
37    /m/0ggx5q   Dance music
38    /m/025td0t   Background music
39    /m/0l14qv   Synthesizer
40    /m/06j6l   Rhythm and blues
41    /m/01xqw   Cello
42    /m/01bjv   Bus
43    /m/07xzm   Ukulele
44    /m/0m0jc   Electronica
45    /m/01v1d8   Sampler
46    /m/0k5j   Aircraft
47    /m/0dwsp   Marimba- xylophone
48    /m/01lyv   Country
49    /t/dd00035   Exciting music
50    /m/01j3sz   Laughter