Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,52 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[](https://huggingface.co/datasets/CSALT/deepfake_detection_dataset_urdu)
|
| 2 |
+
|
| 3 |
+
# Deepfake Defense: Constructing and Evaluating a Specialized Urdu Deepfake Audio Dataset
|
| 4 |
+
|
| 5 |
+
This repository contains the Urdu Deepfake Audio Dataset introduced in the ACL 2024 paper "Deepfake Defense: Constructing and Evaluating a Specialized Urdu Deepfake Audio Dataset".
|
| 6 |
+
|
| 7 |
+
The dataset focuses on two spoofing attacks – Tacotron and VITS TTS – and includes bonafide audio samples for comparison. The dataset construction ensures phonemic cover and balance, making it suitable for training deepfake detection models in Urdu.
|
| 8 |
+
|
| 9 |
+
### Dataset Statistics
|
| 10 |
+
|
| 11 |
+
The dataset includes the following four parts:
|
| 12 |
+
|
| 13 |
+
1. Bonafide Part 1
|
| 14 |
+
2. Bonafide Part 2
|
| 15 |
+
3. Tacotron
|
| 16 |
+
4. VITS TTS
|
| 17 |
+
|
| 18 |
+
The statistics for each part are as follows:
|
| 19 |
+
|
| 20 |
+
| **Metric** | **Bonafide Part 1** | **Bonafide Part 2** | **Tacotron** | **VITS TTS** |
|
| 21 |
+
|------------------------------|---------------------|---------------------|--------------|--------------|
|
| 22 |
+
| **Total Duration (mins)** | 1,302.66 | 1,271.65 | 1,061.96 | 1,340.79 |
|
| 23 |
+
| **Max Sample Length (mins)** | 112.42 | 120.75 | 80.34 | 111.01 |
|
| 24 |
+
| **Min Sample Length (mins)** | 61.73 | 56.45 | 44.64 | 65.53 |
|
| 25 |
+
| **Avg Sample Length (mins)** | 76.63 | 74.80 | 62.47 | 78.87 |
|
| 26 |
+
| **Files per Speaker** | 708 audio files | 495 audio files | 495 audio files | 495 audio files |
|
| 27 |
+
|
| 28 |
+
## Structure
|
| 29 |
+
|
| 30 |
+
The dataset is organized into folders, each containing audio files for the respective parts mentioned above. Each folder is named according to its part (e.g., `Bonafide_Part1`, `Tacotron`, etc.).
|
| 31 |
+
|
| 32 |
+
## Usage
|
| 33 |
+
|
| 34 |
+
The dataset is available on Huggingface through the following link:
|
| 35 |
+
- Huggingface Dataset: https://huggingface.co/datasets/CSALT/deepfake_detection_dataset_urdu
|
| 36 |
+
The code for this project is on Github:
|
| 37 |
+
- https://github.com/CSALT-LUMS/urdu-deepfake-dataset
|
| 38 |
+
|
| 39 |
+
## Citation
|
| 40 |
+
```
|
| 41 |
+
@inproceedings{sheza-etal-2024-deepfake,
|
| 42 |
+
title = "Deepfake Defense: Constructing and Evaluating a Specialized Urdu Deepfake Audio Dataset",
|
| 43 |
+
author = "Sheza Munir, Wassay Sajjad, Mukeet Raza, Emaan Mujahid Abbas, Abdul Hameed Azeemi, Ihsan Ayyub Qazi, and Agha Ali Raza",
|
| 44 |
+
booktitle = "Findings of the Association for Computational Linguistics: ACL 2024",
|
| 45 |
+
year = "2024",
|
| 46 |
+
publisher = "Association for Computational Linguistics",
|
| 47 |
+
}
|
| 48 |
+
```
|
| 49 |
+
|
| 50 |
+
## Legal
|
| 51 |
+
|
| 52 |
+
CC BY-NC 4.0 license for the data hosted on HuggingFace and Google Drive.
|