audio
audioduration (s)
0.06
4.01
slice_file_name
stringlengths
14
18
fsID
int64
2.94k
210k
start
float64
0
600
end
float64
0.11
604
salience
int64
1
2
fold
int64
1
10
classID
int64
0
9
class
stringclasses
10 values
100032-3-0-0.wav
100,032
0
0.317551
1
5
3
dog_bark
100263-2-0-117.wav
100,263
58.5
62.5
1
5
2
children_playing
100263-2-0-121.wav
100,263
60.5
64.5
1
5
2
children_playing
100263-2-0-126.wav
100,263
63
67
1
5
2
children_playing
100263-2-0-137.wav
100,263
68.5
72.5
1
5
2
children_playing
100263-2-0-143.wav
100,263
71.5
75.5
1
5
2
children_playing
100263-2-0-161.wav
100,263
80.5
84.5
1
5
2
children_playing
100263-2-0-3.wav
100,263
1.5
5.5
1
5
2
children_playing
100263-2-0-36.wav
100,263
18
22
1
5
2
children_playing
100648-1-0-0.wav
100,648
4.823402
5.471927
2
10
1
car_horn
100648-1-1-0.wav
100,648
8.998279
10.052132
2
10
1
car_horn
100648-1-2-0.wav
100,648
16.699509
17.104837
2
10
1
car_horn
100648-1-3-0.wav
100,648
17.631764
19.253075
2
10
1
car_horn
100648-1-4-0.wav
100,648
25.332994
27.197502
2
10
1
car_horn
100652-3-0-0.wav
100,652
0
4
1
2
3
dog_bark
100652-3-0-1.wav
100,652
0.5
4.5
1
2
3
dog_bark
100652-3-0-2.wav
100,652
1
5
1
2
3
dog_bark
100652-3-0-3.wav
100,652
1.5
5.5
1
2
3
dog_bark
100795-3-0-0.wav
100,795
0.19179
4.19179
1
10
3
dog_bark
100795-3-1-0.wav
100,795
13.059155
17.059155
1
10
3
dog_bark
100795-3-1-1.wav
100,795
13.559155
17.559155
1
10
3
dog_bark
100795-3-1-2.wav
100,795
14.059155
18.059155
1
10
3
dog_bark
100852-0-0-0.wav
100,852
0
4
1
5
0
air_conditioner
100852-0-0-1.wav
100,852
0.5
4.5
1
5
0
air_conditioner
100852-0-0-10.wav
100,852
5
9
1
5
0
air_conditioner
100852-0-0-11.wav
100,852
5.5
9.5
1
5
0
air_conditioner
100852-0-0-12.wav
100,852
6
10
1
5
0
air_conditioner
100852-0-0-13.wav
100,852
6.5
10.5
1
5
0
air_conditioner
100852-0-0-14.wav
100,852
7
11
1
5
0
air_conditioner
100852-0-0-15.wav
100,852
7.5
11.5
1
5
0
air_conditioner
100852-0-0-16.wav
100,852
8
12
1
5
0
air_conditioner
100852-0-0-17.wav
100,852
8.5
12.5
1
5
0
air_conditioner
100852-0-0-18.wav
100,852
9
13
1
5
0
air_conditioner
100852-0-0-19.wav
100,852
9.5
13.5
1
5
0
air_conditioner
100852-0-0-2.wav
100,852
1
5
1
5
0
air_conditioner
100852-0-0-20.wav
100,852
10
14
1
5
0
air_conditioner
100852-0-0-21.wav
100,852
10.5
14.5
1
5
0
air_conditioner
100852-0-0-22.wav
100,852
11
15
1
5
0
air_conditioner
100852-0-0-23.wav
100,852
11.5
15.5
1
5
0
air_conditioner
100852-0-0-24.wav
100,852
12
16
1
5
0
air_conditioner
100852-0-0-25.wav
100,852
12.5
16.5
1
5
0
air_conditioner
100852-0-0-26.wav
100,852
13
17
1
5
0
air_conditioner
100852-0-0-27.wav
100,852
13.5
17.5
1
5
0
air_conditioner
100852-0-0-28.wav
100,852
14
18
1
5
0
air_conditioner
100852-0-0-29.wav
100,852
14.5
18.5
1
5
0
air_conditioner
100852-0-0-3.wav
100,852
1.5
5.5
1
5
0
air_conditioner
100852-0-0-30.wav
100,852
15
19
1
5
0
air_conditioner
100852-0-0-4.wav
100,852
2
6
1
5
0
air_conditioner
100852-0-0-5.wav
100,852
2.5
6.5
1
5
0
air_conditioner
100852-0-0-6.wav
100,852
3
7
1
5
0
air_conditioner
100852-0-0-7.wav
100,852
3.5
7.5
1
5
0
air_conditioner
100852-0-0-8.wav
100,852
4
8
1
5
0
air_conditioner
100852-0-0-9.wav
100,852
4.5
8.5
1
5
0
air_conditioner
101281-3-0-0.wav
101,281
0.341333
4.341333
2
6
3
dog_bark
101281-3-0-14.wav
101,281
7.341333
11.341333
2
6
3
dog_bark
101281-3-0-5.wav
101,281
2.841333
6.841333
2
6
3
dog_bark
101382-2-0-10.wav
101,382
5
9
1
10
2
children_playing
101382-2-0-12.wav
101,382
6
10
1
10
2
children_playing
101382-2-0-20.wav
101,382
10
14
1
10
2
children_playing
101382-2-0-21.wav
101,382
10.5
14.5
1
10
2
children_playing
101382-2-0-29.wav
101,382
14.5
18.5
1
10
2
children_playing
101382-2-0-33.wav
101,382
16.5
20.5
1
10
2
children_playing
101382-2-0-42.wav
101,382
21
25
1
10
2
children_playing
101382-2-0-45.wav
101,382
22.5
26.5
1
10
2
children_playing
101415-3-0-2.wav
101,415
1
5
1
1
3
dog_bark
101415-3-0-3.wav
101,415
1.5
5.5
1
1
3
dog_bark
101415-3-0-8.wav
101,415
4
8
1
1
3
dog_bark
101729-0-0-1.wav
101,729
0.5
4.5
2
9
0
air_conditioner
101729-0-0-11.wav
101,729
5.5
9.5
2
9
0
air_conditioner
101729-0-0-12.wav
101,729
6
10
2
9
0
air_conditioner
101729-0-0-13.wav
101,729
6.5
10.5
2
9
0
air_conditioner
101729-0-0-14.wav
101,729
7
11
2
9
0
air_conditioner
101729-0-0-16.wav
101,729
8
12
2
9
0
air_conditioner
101729-0-0-17.wav
101,729
8.5
12.5
2
9
0
air_conditioner
101729-0-0-18.wav
101,729
9
13
2
9
0
air_conditioner
101729-0-0-19.wav
101,729
9.5
13.5
2
9
0
air_conditioner
101729-0-0-21.wav
101,729
10.5
14.5
2
9
0
air_conditioner
101729-0-0-22.wav
101,729
11
15
2
9
0
air_conditioner
101729-0-0-23.wav
101,729
11.5
15.5
2
9
0
air_conditioner
101729-0-0-24.wav
101,729
12
16
2
9
0
air_conditioner
101729-0-0-26.wav
101,729
13
17
2
9
0
air_conditioner
101729-0-0-28.wav
101,729
14
18
2
9
0
air_conditioner
101729-0-0-29.wav
101,729
14.5
18.5
2
9
0
air_conditioner
101729-0-0-3.wav
101,729
1.5
5.5
2
9
0
air_conditioner
101729-0-0-32.wav
101,729
16
20
2
9
0
air_conditioner
101729-0-0-33.wav
101,729
16.5
20.5
2
9
0
air_conditioner
101729-0-0-36.wav
101,729
18
22
2
9
0
air_conditioner
101729-0-0-37.wav
101,729
18.5
22.5
2
9
0
air_conditioner
101729-0-0-38.wav
101,729
19
23
2
9
0
air_conditioner
101729-0-0-39.wav
101,729
19.5
23.5
2
9
0
air_conditioner
101729-0-0-4.wav
101,729
2
6
2
9
0
air_conditioner
101729-0-0-40.wav
101,729
20
24
2
9
0
air_conditioner
101729-0-0-6.wav
101,729
3
7
2
9
0
air_conditioner
101729-0-0-9.wav
101,729
4.5
8.5
2
9
0
air_conditioner
101848-9-0-0.wav
101,848
0
4
1
7
9
street_music
101848-9-0-1.wav
101,848
0.5
4.5
1
7
9
street_music
101848-9-0-2.wav
101,848
1
5
1
7
9
street_music
101848-9-0-3.wav
101,848
1.5
5.5
1
7
9
street_music
101848-9-0-8.wav
101,848
4
8
1
7
9
street_music
101848-9-0-9.wav
101,848
4.5
8.5
1
7
9
street_music

(card and dataset copied from https://www.kaggle.com/datasets/chrisfilo/urbansound8k)

This dataset contains 8732 labeled sound excerpts (<=4s) of urban sounds from 10 classes: air_conditioner, car_horn, children_playing, dog_bark, drilling, enginge_idling, gun_shot, jackhammer, siren, and street_music. The classes are drawn from the urban sound taxonomy. For a detailed description of the dataset and how it was compiled please refer to our paper.All excerpts are taken from field recordings uploaded to www.freesound.org. The files are pre-sorted into ten folds (folders named fold1-fold10) to help in the reproduction of and comparison with the automatic classification results reported in the article above.

In addition to the sound excerpts, a CSV file containing metadata about each excerpt is also provided.

AUDIO FILES INCLUDED

8732 audio files of urban sounds (see description above) in WAV format. The sampling rate, bit depth, and number of channels are the same as those of the original file uploaded to Freesound (and hence may vary from file to file).

META-DATA FILES INCLUDED

UrbanSound8k.csv

This file contains meta-data information about every audio file in the dataset. This includes:

  • slice_file_name: The name of the audio file. The name takes the following format: [fsID]-[classID]-[occurrenceID]-[sliceID].wav, where: [fsID] = the Freesound ID of the recording from which this excerpt (slice) is taken [classID] = a numeric identifier of the sound class (see description of classID below for further details) [occurrenceID] = a numeric identifier to distinguish different occurrences of the sound within the original recording [sliceID] = a numeric identifier to distinguish different slices taken from the same occurrence

  • fsID: The Freesound ID of the recording from which this excerpt (slice) is taken

  • start The start time of the slice in the original Freesound recording

  • end: The end time of slice in the original Freesound recording

  • salience: A (subjective) salience rating of the sound. 1 = foreground, 2 = background.

  • fold: The fold number (1-10) to which this file has been allocated.

  • classID: A numeric identifier of the sound class: 0 = air_conditioner 1 = car_horn 2 = children_playing 3 = dog_bark 4 = drilling 5 = engine_idling 6 = gun_shot 7 = jackhammer 8 = siren 9 = street_music

  • class: The class name: air_conditioner, car_horn, children_playing, dog_bark, drilling, engine_idling, gun_shot, jackhammer, siren, street_music.

BEFORE YOU DOWNLOAD: AVOID COMMON PITFALLS!

Since releasing the dataset we have noticed a couple of common mistakes that could invalidate your results, potentially leading to manuscripts being rejected or the publication of incorrect results. To avoid this, please read the following carefully:

  1. Don't reshuffle the data! Use the predefined 10 folds and perform 10-fold (not 5-fold) cross validation The experiments conducted by vast majority of publications using UrbanSound8K (by ourselves and others) evaluate classification models via 10-fold cross validation using the predefined splits*. We strongly recommend following this procedure.

Why? If you reshuffle the data (e.g. combine the data from all folds and generate a random train/test split) you will be incorrectly placing related samples in both the train and test sets, leading to inflated scores that don't represent your model's performance on unseen data. Put simply, your results will be wrong. Your results will NOT be comparable to previous results in the literature, meaning any claims to an improvement on previous research will be invalid. Even if you don't reshuffle the data, evaluating using different splits (e.g. 5-fold cross validation) will mean your results are not comparable to previous research.

  1. Don't evaluate just on one split! Use 10-fold (not 5-fold) cross validation and average the scores We have seen reports that only provide results for a single train/test split, e.g. train on folds 1-9, test on fold 10 and report a single accuracy score. We strongly advise against this. Instead, perform 10-fold cross validation using the provided folds and report the average score.

Why? Not all the splits are as "easy". That is, models tend to obtain much higher scores when trained on folds 1-9 and tested on fold 10, compared to (e.g.) training on folds 2-10 and testing on fold 1. For this reason, it is important to evaluate your model on each of the 10 splits and report the average accuracy. Again, your results will NOT be comparable to previous results in the literature.

Acknowledgements

We kindly request that articles and other works in which this dataset is used cite the following paper:

J. Salamon, C. Jacoby and J. P. Bello, "A Dataset and Taxonomy for Urban Sound Research", 22nd ACM International Conference on Multimedia, Orlando USA, Nov. 2014.

More information at https://urbansounddataset.weebly.com/urbansound8k.html

Downloads last month
700
Edit dataset card