The Audio, Speech, and Vision Processing Lab - Emotional Sound Database (ASVP - ESD)
ABOUT
The Audio, Speech, and Vision Processing Lab - Emotional Sound Database (ASVP - ESD)
was created by School of Electronic and Information Engineering, South China University of Technology.
CHOSEN EMOTIONS
13 emotions were chosen:
boredom,sigh
neutral,calm
happy,laugh,gaggle
sad,cry
angry,grunt,frustration
fearful,scream,panic
disgust,dislike,contempt
surprised,gasp,amazed
excited
pleasure
pain,groan
disappointment,disapproval
breath
ORGANISING THE DATABASE
Speech Statistic
Duration Statisitc
Duration
Num. of Clips
2,150
Total Duration
13347.835 seconds = 222.464 minutes = 3.708 hours
Max Dur
32.235 seconds
Min Dur
0.287 seconds
Mean Dur
6.208 seconds
Std. Dur
3.839 seconds
Num. of Clips > 30 seconds
1
Emotion
Num. of Clips
01: boredom, sigh
81
02: neutral, calm
657
03: happy, laugh, gaggle
154
04: sad, cry
268
05: angry, grunt, frustration
385
06: fearful, scream, panic
63
07: disgust, dislike, contempt
90
08: surprised, hasp, amazed
144
09: excited
136
10: pleasure
15
11: pain, groan
25
12: disappointmrnt, disapproval
132
13: breath
0
Emotion Intensity
Num. of Clips
01: normal
1,783
02: high
367
Gender
Num. of Clips
01: male
1,224
02: female
926
Age Range
Num. of Clips
01: >65
65
02: 20~65
1,914
03: 3<20
80
04: <3
91
Language
Num. of Clips
01: Mandarin
937
02: English
621
03: French
175
04: Others
417
Non-Speech Statistic
Duration Statisitc
Duration
Num. of Clips
5,484
Total Duration
14438.117 seconds = 240.635 minutes = 4.011 hours
Max Dur
25.810 seconds
Min Dur
0.141 seconds
Mean Dur
2.633 seconds
Std. Dur
2.720 seconds
Num. of Clips > 30 seconds
0
Emotion
Num. of Clips
01: boredom, sigh
392
02: neutral, calm
253
03: happy, laugh, gaggle
878
04: sad, cry
383
05: angry, grunt, frustration
339
06: fearful, scream, panic
799
07: disgust, dislike, contempt
473
08: surprised, hasp, amazed
808
09: excited
109
10: pleasure
273
11: pain, groan
706
12: disappointmrnt, disapproval
70
13: breath
1
Emotion Intensity
Num. of Clips
01: normal
4,693
02: high
791
Gender
Num. of Clips
01: male
2,919
02: female
2,565
Age Range
Num. of Clips
01: >65
73
02: 20~65
5,224
03: 3<20
100
04: <3
87
Language
Num. of Clips
01: Mandarin
512
02: English
3,258
03: French
109
04: Others
1,605
References
Dejoli Tientcheu Touko Landry, Qianhua He, Haikang Yan and Yanxiong Li. (2020). ASVP-ESD:A dataset and its benchmark for emotion recognition using both speech and non-speech utterances. Global Scientific Journals, 8(6), 1793-1798.