Dataset Viewer
Auto-converted to Parquet
audio
audio
filename
string
category
string
timestamp
string
label
int64
manipulation_type
string
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_21
real
0.00-7.92-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_30
real
0.00-7.92-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_12
real
0.00-7.36-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_8
real
0.00-7.08-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_44
real
0.00-7.04-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_10
real
0.00-7.00-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_03_baum_64kb_19
real
0.00-6.96-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_7
real
0.00-6.84-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_13
real
0.00-6.80-T
1
real
dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_46
real
0.00-6.80-T
1
real
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_21
edit
0.00-4.89-T/4.89-5.19-F/5.19-8.01-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_30
edit
0.00-1.09-T/1.09-1.54-F/1.54-7.74-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_12
edit
0.00-4.93-T/4.93-5.64-F/5.64-7.47-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_8
edit
0.00-0.57-T/0.57-1.08-F/1.08-7.00-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_44
edit
0.00-5.92-T/5.92-6.49-F/6.49-6.91-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_10
edit
0.00-1.26-T/1.26-1.74-F/1.74-6.89-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_03_baum_64kb_19
edit
0.00-0.61-T/0.61-0.94-F/0.94-7.02-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_7
edit
0.00-3.34-T/3.34-3.45-F/3.45-6.71-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_13
edit
0.00-5.86-T/5.86-6.31-F/6.31-6.95-T
0
edit
dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_46
edit
0.00-0.11-T/0.11-0.62-F/0.62-6.77-T
0
edit
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_21
cut_paste
0.00-4.89-T/4.89-5.19-F/5.19-8.01-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_30
cut_paste
0.00-1.09-T/1.09-1.54-F/1.54-7.74-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_12
cut_paste
0.00-4.93-T/4.93-5.64-F/5.64-7.47-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_8
cut_paste
0.00-0.57-T/0.57-1.08-F/1.08-7.00-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_44
cut_paste
0.00-5.92-T/5.92-6.49-F/6.49-6.91-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_10
cut_paste
0.00-1.26-T/1.26-1.74-F/1.74-6.89-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_03_baum_64kb_19
cut_paste
0.00-0.61-T/0.61-0.94-F/0.94-7.02-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_7
cut_paste
0.00-3.34-T/3.34-3.45-F/3.45-6.71-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_13
cut_paste
0.00-5.86-T/5.86-6.31-F/6.31-6.95-T
0
cut_paste
dev_cut_paste_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_01_baum_64kb_46
cut_paste
0.00-0.11-T/0.11-0.62-F/0.62-6.77-T
0
cut_paste

SINE Dataset

Overview

The Speech INfilling Edit (SINE) dataset is a comprehensive collection for speech deepfake detection and audio authenticity verification. This dataset contains ~87GB of audio data distributed across 32 splits, featuring both authentic and synthetically manipulated speech samples.

Dataset Statistics

  • Total Size: ~87GB
  • Number of Splits: 32 (split-0.tar.gz to split-31.tar.gz)
  • Audio Format: WAV files
  • Source: Speech edited from LibriLight dataset with transcripts obtained from LibriHeavy

Audio Statistics

Audio Types Subsets # of Samples # of Speakers Durations (h) Audio Lengths (s)
min max
Real/Resyn train 26,547 70 51.82 6.00 8.00
val 8,676 100 16.98 6.00 8.00
test 8,494 900 16.60 6.00 8.00
Infill/CaP train 26,546 70 51.98 5.40 9.08
val 8,686 100 16.99 5.45 8.76
test 8,493 903 16.64 5.49 8.85

Data Structure

Each split (e.g., split-0/) contains:

split-X/
β”œβ”€β”€ combine/                    # Directory containing all audio files (~11,076 files)
β”‚   β”œβ”€β”€ dev_real_medium-*.wav          # Authentic audio samples
β”‚   β”œβ”€β”€ dev_edit_medium-*.wav          # Edited audio samples
β”‚   β”œβ”€β”€ dev_cut_paste_medium-*.wav     # Cut-and-paste manipulated samples
β”‚   └── dev_resyn_medium-*.wav         # Resynthesized audio samples
β”œβ”€β”€ medium_real.txt             # Labels for authentic audio (2,769 entries)
β”œβ”€β”€ medium_edit.txt             # Labels for edited audio (2,769 entries)
β”œβ”€β”€ medium_cut_paste.txt        # Labels for cut-paste audio (2,769 entries)
└── medium_resyn.txt            # Labels for resynthesized audio (2,769 entries)

Audio Categories

1. Authentic Speech (dev_real_medium-*)

  • Original, unmodified speech recordings from LibriVox audiobooks
  • Labeled as class 1 (authentic)
  • Simple time annotation format: filename start-end-T label

2. Resynthesized Speech (dev_resyn_medium-*)

  • Speech regenerated from mel-spectrogram using HiFi-GAN vocoder
  • Labeled as class 1 (authentic)
  • Simple time annotation format

3. Edited Speech (dev_edit_medium-*)

  • Audio samples with artificial modifications/edits
  • Labeled as class 0 (manipulated)
  • Complex time annotation with T/F segments indicating real/fake portions

4. Cut-and-Paste Speech (dev_cut_paste_medium-*)

  • Audio created by cutting and pasting segments from different sources
  • Labeled as class 0 (manipulated)
  • Complex time annotation showing spliced segments

Label Format

Simple Format (Real/Resyn)

filename start_time-end_time-T label

Example:

dev_real_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_21 0.00-7.92-T 1

Complex Format (Edit/Cut-Paste)

filename time_segment1-T/time_segment2-F/time_segment3-T label

Example:

dev_edit_medium-100-emerald_city_librivox_64kb_mp3-emeraldcity_02_baum_64kb_21 0.00-4.89-T/4.89-5.19-F/5.19-8.01-T 0

Where:

  • T = True/Authentic segment
  • F = False/Manipulated segment
  • label: 1 = Authentic, 0 = Manipulated

Applications

This dataset is suitable for:

  • Speech Deepfake Detection: Binary classification of authentic vs. manipulated speech
  • Temporal Localization: Identifying specific time segments that contain manipulations
  • Manipulation Type Classification: Distinguishing between different types of audio manipulation
  • Robustness Testing: Evaluating detection systems across various manipulation techniques

Citation

This is a joint work done by NVIDIA and National Taiwan University. If you use this dataset, please cite:

@inproceedings{huang2024detecting,
  title={Detecting the Undetectable: Assessing the Efficacy of Current Spoof Detection Methods Against Seamless Speech Edits},
  author={Huang, Sung-Feng and Kuo, Heng-Cheng and Chen, Zhehuai and Yang, Xuesong and Yang, Chao-Han Huck and Tsao, Yu and Wang, Yu-Chiang Frank and Lee, Hung-yi and Fu, Szu-Wei},
  booktitle={2024 IEEE Spoken Language Technology Workshop (SLT)},
  pages={652--659},
  year={2024},
  organization={IEEE}
}

License

This dataset is released under the Apache 2.0 License.


Note: This dataset is intended for research purposes in speech authenticity verification and deepfake detection. Please use responsibly and in accordance with applicable laws and regulations.

Downloads last month
47