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/07q0yl5   Snort
2    /m/07q4ntr   Bellow
3    /m/016622   Tubular bells
4    /m/02mfyn   Car alarm
5    /m/07rgkc5   Static
6    /m/02yds9   Purr
7    /m/01y3hg   Smoke detector
8    /m/026fgl   Wind chime
9    /m/07qv_x_   Shuffle
10    /m/046dlr   Alarm clock
11    /m/07q5rw0   Neigh- whinny
12    /m/0xzly   Maraca
13    /m/07r5c2p   Caw
14    /m/07ptzwd   Pump
15    /m/0l7xg   Gears
16    /m/0c2wf   Typewriter
17    /m/07kc_   Theremin
18    /m/09d5_   Owl
19    /m/07p7b8y   Fill (with liquid)
20    /m/06hps   Rodents
21    /t/dd00006   Synthetic singing
22    /m/0l156k   Whistle
23    /m/0150b9   Change ringing
24    /m/073cg4   Cap gun
25    /t/dd00135   Children shouting
26    /m/07rpkh9   Moo
27    /m/02p01q   Filing
28    /m/02bm9n   Ratchet
29    /m/07s2xch   Groan
30    /m/07pp8cl   Telephone bell ringing
31    /m/07r4gkf   Patter
32    /m/01h82_   Engine knocking
33    /m/01x3z   Clock
34    /m/07brj   Tambourine
35    /m/0b9m1   Harmonic
36    /m/07qv_d5   Toot
37    /m/04229   Jet engine
38    /m/04zmvq   Train whistle
39    /m/07qc9xj   Clicking
40    /m/07phxs1   Ding
41    /m/07r5v4s   Drip
42    /m/07q7njn   Chink - clink
43    /m/07pp_mv   Alarm
44    /m/0ghcn6   Growling
45    /m/0l15bq   Clapping
46    /m/0cdnk   Roaring cats (lions tigers)
47    /m/07sr1lc   Yell
48    /m/07pxg6y   Eruption
49    /m/07qwdck   Ping
50    /m/01b82r   Sawing