segment_id
stringlengths 10
10
| audio
audioduration (s) 0
38.8
| dialect
stringclasses 5
values | domain
stringclasses 4
values | audio_duration
float64 0
38.8
|
|---|---|---|---|---|
XeNHoyXIbC
|
Egyptian Arabic
|
Theatre
| 4.564
|
|
1LjVS0Nz8E
|
Gulf Arabic
|
Radio
| 5.484
|
|
w5YmAjSJ4m
|
Egyptian Arabic
|
TV Dramas
| 5
|
|
SVLVEwxoUw
|
Maghrebi Arabic
|
TED Talks
| 13.575
|
|
mmdB55Qd9N
|
Maghrebi Arabic
|
Radio
| 9.506
|
|
GIFtPhQcYz
|
Gulf Arabic
|
TED Talks
| 6.876
|
|
iEFeuxJftq
|
Gulf Arabic
|
TV Dramas
| 5
|
|
VrBqk9NqZj
|
Gulf Arabic
|
TED Talks
| 12.273
|
|
fXBlzqPDoM
|
Levantine Arabic
|
TED Talks
| 9.859
|
|
pQI1G59vI1
|
Maghrebi Arabic
|
Theatre
| 15.825
|
|
d8CSJNNBHM
|
Egyptian Arabic
|
Theatre
| 3.202
|
|
6NzPoXlBGD
|
Levantine Arabic
|
TED Talks
| 6.438
|
|
Mxgx7YOwso
|
Egyptian Arabic
|
TED Talks
| 4.314
|
|
9mdYYlaq0Q
|
Egyptian Arabic
|
TV Dramas
| 7
|
|
xIfhfkkBRX
|
Egyptian Arabic
|
TV Dramas
| 5
|
|
bSlqymgSYe
|
Levantine Arabic
|
Radio
| 9.633
|
|
7lpp4sNA1l
|
Levantine Arabic
|
TED Talks
| 7.674
|
|
htdVLxNrOI
|
Modern Standard Arabic
|
Radio
| 8.422
|
|
klDPGDo3s7
|
Egyptian Arabic
|
TED Talks
| 11.404
|
|
c0t7KzKwFv
|
Gulf Arabic
|
Radio
| 7.756
|
|
dKsOiBsx9l
|
Levantine Arabic
|
Radio
| 9.769
|
|
ONGfZmjjbL
|
Levantine Arabic
|
Theatre
| 8.642
|
|
B4sJUS9ffQ
|
Maghrebi Arabic
|
Radio
| 17.213
|
|
DXyEyBHrEc
|
Levantine Arabic
|
TED Talks
| 19.482
|
|
DuTvzB42mq
|
Gulf Arabic
|
TED Talks
| 6.352
|
|
KSoSksBhX3
|
Egyptian Arabic
|
TED Talks
| 8.357
|
|
jLRiLXVDpw
|
Egyptian Arabic
|
TED Talks
| 9.246
|
|
NMznNTc6kf
|
Egyptian Arabic
|
TED Talks
| 9.332
|
|
pfZcaxYFM0
|
Modern Standard Arabic
|
Radio
| 10.651
|
|
igQ6kgMX5t
|
Levantine Arabic
|
Theatre
| 6.163
|
|
6fdm0WcmhO
|
Egyptian Arabic
|
Radio
| 6.618
|
|
goebW2zy5I
|
Gulf Arabic
|
TED Talks
| 15.381
|
|
0mg8lfyBi3
|
Levantine Arabic
|
TV Dramas
| 5
|
|
bfShjFp3tK
|
Egyptian Arabic
|
TED Talks
| 4.314
|
|
hV4AoWEi8a
|
Levantine Arabic
|
Radio
| 6.63
|
|
Fhg63lgNCY
|
Egyptian Arabic
|
Theatre
| 5.987
|
|
tHD7huGjhD
|
Egyptian Arabic
|
TED Talks
| 8.259
|
|
GB3DLzAqol
|
Egyptian Arabic
|
TED Talks
| 11.09
|
|
c7lij5QA7r
|
Modern Standard Arabic
|
Theatre
| 11.688
|
|
1ki046Ba80
|
Maghrebi Arabic
|
Radio
| 12.984
|
|
PhwIK8R7ay
|
Gulf Arabic
|
TV Dramas
| 5
|
|
pplwSJMcCg
|
Levantine Arabic
|
TED Talks
| 7.668
|
|
iGCZWXFyqo
|
Levantine Arabic
|
TV Dramas
| 5
|
|
ob70nVCLuk
|
Egyptian Arabic
|
TV Dramas
| 5
|
|
Do7zELJI8c
|
Egyptian Arabic
|
TED Talks
| 7.125
|
|
7yFNsqkQKi
|
Maghrebi Arabic
|
TED Talks
| 13.176
|
|
knAEA26xdJ
|
Gulf Arabic
|
Radio
| 9.264
|
|
giqHJfauiS
|
Egyptian Arabic
|
TED Talks
| 10.049
|
|
WzmV5FhBwk
|
Egyptian Arabic
|
TED Talks
| 7.747
|
|
8PtybL8J7k
|
Maghrebi Arabic
|
Radio
| 19.64
|
|
ClRK52z0Ra
|
Levantine Arabic
|
Theatre
| 10.847
|
|
bPbL0PZtol
|
Egyptian Arabic
|
TED Talks
| 4.468
|
|
I6HFcPzxkV
|
Maghrebi Arabic
|
Radio
| 8.891
|
|
wG96cUmdZH
|
Egyptian Arabic
|
TED Talks
| 6.59
|
|
TE3dnbmvsi
|
Maghrebi Arabic
|
Radio
| 7.336
|
|
xjUJq6oTJO
|
Levantine Arabic
|
TV Dramas
| 5
|
|
7x450E2Kpn
|
Egyptian Arabic
|
Theatre
| 5.92
|
|
0M7cC6ZzfY
|
Gulf Arabic
|
TED Talks
| 8.975
|
|
AyUNJk6pRW
|
Gulf Arabic
|
Theatre
| 7.52
|
|
iwUIrusWCk
|
Levantine Arabic
|
Theatre
| 4.603
|
|
JjQwrPyUKo
|
Gulf Arabic
|
Radio
| 14.116
|
|
kd4La9xx1J
|
Maghrebi Arabic
|
Radio
| 11.325
|
|
Zh6ss2jsYe
|
Modern Standard Arabic
|
Theatre
| 8.815
|
|
EloUZbaBNL
|
Maghrebi Arabic
|
Theatre
| 4.951
|
|
l9KaR2aCG8
|
Gulf Arabic
|
TED Talks
| 15.864
|
|
6mlMt049Mx
|
Levantine Arabic
|
Theatre
| 6.525
|
|
N4k5Uzshvc
|
Maghrebi Arabic
|
Radio
| 9.87
|
|
JRnOqbFFeY
|
Levantine Arabic
|
TV Dramas
| 5
|
|
TdcqiYOtkz
|
Maghrebi Arabic
|
Radio
| 16.911
|
|
PpgeGd7wNU
|
Modern Standard Arabic
|
Theatre
| 9.274
|
|
JVzssSFVZb
|
Maghrebi Arabic
|
Radio
| 7.366
|
|
8QLV86QKhi
|
Maghrebi Arabic
|
Radio
| 6.151
|
|
sjC8UgmJH1
|
Gulf Arabic
|
TV Dramas
| 5
|
|
Vhpp1Jy8kc
|
Egyptian Arabic
|
TED Talks
| 10.178
|
|
AegF8CIw2O
|
Egyptian Arabic
|
TED Talks
| 10.414
|
|
MAaM2BFdrt
|
Modern Standard Arabic
|
Theatre
| 16.411
|
|
ipyR1ElKqH
|
Maghrebi Arabic
|
Radio
| 14.118
|
|
5N4E6gkzcL
|
Gulf Arabic
|
TV Dramas
| 5
|
|
z7rwJ4uaze
|
Levantine Arabic
|
Theatre
| 9.959
|
|
OGQAwEaexk
|
Gulf Arabic
|
TV Dramas
| 5
|
|
FiW7MFi85i
|
Modern Standard Arabic
|
Theatre
| 6.729
|
|
YoUrGqAgvc
|
Egyptian Arabic
|
TV Dramas
| 5
|
|
4uXdAmFkUY
|
Maghrebi Arabic
|
Radio
| 9.052
|
|
XF5AWKA9LW
|
Gulf Arabic
|
TV Dramas
| 5
|
|
85nChWchgS
|
Gulf Arabic
|
TED Talks
| 9.196
|
|
Hc9I6jQFjo
|
Gulf Arabic
|
TV Dramas
| 5
|
|
TNvlxnVJkc
|
Modern Standard Arabic
|
Theatre
| 14.159
|
|
UwvCVliSCr
|
Maghrebi Arabic
|
Theatre
| 6.987
|
|
Vq3LikmXg9
|
Gulf Arabic
|
TED Talks
| 8.789
|
|
ICQYTRccTH
|
Modern Standard Arabic
|
Theatre
| 7.54
|
|
2i242w4xvp
|
Gulf Arabic
|
TV Dramas
| 5
|
|
GLOJjP3j4P
|
Levantine Arabic
|
TED Talks
| 10.493
|
|
IED9h5vC9D
|
Egyptian Arabic
|
TV Dramas
| 7
|
|
91A7cwItRe
|
Modern Standard Arabic
|
TED Talks
| 8.929
|
|
D9bSIxLWTT
|
Egyptian Arabic
|
TED Talks
| 11.465
|
|
vuvfoLdfzG
|
Maghrebi Arabic
|
Radio
| 8.022
|
|
FPCb5cDAUo
|
Gulf Arabic
|
Radio
| 9.025
|
|
aPXDrIbv9E
|
Levantine Arabic
|
TED Talks
| 6.855
|
|
GwfCaPLO2H
|
Maghrebi Arabic
|
Radio
| 15.351
|
|
didtj0snvw
|
Levantine Arabic
|
Theatre
| 5.233
|
Dataset Overview
MADIS-5 (Multi-domain Arabic Dialect Identification in Speech) is a manually curated dataset designed to facilitate evaluation of cross-domain robustness of Arabic Dialect Identification (ADI) systems. This dataset provides a comprehensive benchmark for testing out-of-domain generalization across different speech domains with diverse recording conditions and speaking styles.
Dataset Statistics
- Total Duration: ~12 hours of speech
- Total Utterances: 4,854 utterances
- Languages/Dialects: 5 major Arabic varieties
- Modern Standard Arabic (MSA)
- Egyptian Arabic
- Gulf Arabic
- Levantine Arabic
- Maghrebi Arabic
- Domains: 4 different spoken domains
- Collection Period: November 2024 - Feb 2025
Data Sources
Our dataset comprises speech samples from four different public sources, each offering varying degrees of similarity to the TV broadcast domain commonly used in ADI research:
π» Radio Broadcasts
- Source: Local radio stations across the Arab world via radio.garden
- Characteristics: Similar to prior ADI datasets but with more casual, spontaneous speech
- Domain Similarity: High similarity to existing ADI benchmarks
πΊ TV Dramas
- Source: Arabic Spoken Dialects Regional Archive (SARA) on Kaggle
- Characteristics: 5-7 second conversational speech segments
- Domain Similarity: Low similarity with more dialogues
π€ TEDx Talks
- Source: Arabic portion of the TEDx dataset with dialect labels
- Characteristics: Presentations with educational content
- Domain Similarity: Moderate similarity due to topic diversity
π Theater
- Source: YouTube dramatic and comedy plays from various Arab countries
- Characteristics: Theatrical performances spanning different time periods
- Domain Similarity: Low similarity with artistic and performative speech, with occasional poor recording conditions
Annotation Process
Quality Assurance
- Primary Annotator: Native Arabic speaker with PhD in Computational Linguistics and extensive exposure to Arabic language variation
- Verification: Independent verification by a second native Arabic speaker with expertise in Arabic dialects
- Segmentation: Manual segmentation and labeling of all recordings
Inter-Annotator Agreement
- Perfect Agreement: 97.7% of all samples
- Disagreement: 2.3% disagreement on radio broadcast segments (MSA vs. dialect classification)
- Note: The small disagreement reflects the natural continuum between MSA and dialectal Arabic in certain contexts. Final label of segments with disagreement was assigned after a discussion between annotators.
Use Cases
This dataset is ideal for:
- Cross-domain robustness evaluation of Arabic dialect identification systems
- Benchmarking ADI models across diverse speech domains
- Research on domain adaptation in Arabic speech processing
- Development of more robust Arabic dialect classifiers
Dataset Advantages
- Domain Diversity: Four distinct speech domains with varying recording conditions
- Expert Annotation: High-quality labels from linguistic experts
- Cross-domain Focus: Specifically designed to test model robustness beyond single domains
- Real-world Scenarios: Covers authentic speech from various contexts
Citation
If you use this dataset in your research, please cite our paper:
@inproceedings{abdullah2025voice,
title={Voice Conversion Improves Cross-Domain Robustness for Spoken Arabic Dialect Identification},
author={Abdullah, Badr M. and Matthew Baas and Bernd MΓΆbius and Dietrich Klakow},
year={2025},
publisher={Interspeech},
url={arxiv.org/abs/2505.24713}
}
License
Creative Commons Attribution-NonCommercial-NoDerivs 4.0 (CC BY-NC-ND 4.0)
Acknowledgments
We thank the contributors to the source datasets and platforms that made this compilation possible, including radio.garden, SARA archive, and the Multilingual TEDx dataset.
- Downloads last month
- 207