File size: 3,258 Bytes
75f9679 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 | ---
language:
- ja
license: cc-by-nc-4.0
tags:
- audio
- scream
- shout
annotations_creators:
- expert-generated
pretty_name: AGSC
viewer: false
---
Warning: BETA VERSION. NOT FOR PRODUCTION USE
# AGSC: Action Gameplay Social Communication Corpus
## Dataset Description
The **Action Gameplay Social Communication (AGSC) corpus** is a collection of spontaneous voice chat recordings from multi-party online gameplay sessions.
It focuses on naturally occurring screams and shouts, providing resources for the study of non-verbal expressions in social contexts.
- **Size**: ~7.8 GB of audio
- **Modality**: Audio
- **Language**: Japanese
- **Annotations**: Includes Praat TextGrid files, providing utterance-level segments and transcriptions, as well as expression-level annotations of affect bursts including screams, shouts, laughs, and expressive interjections.
---
## Usage
You can load the dataset using the `datasets` library:
```python
from datasets import load_dataset
dataset = load_dataset("hirokisince1998/AGSC")
```
The dataset contains the following components:
- Audio files
- Praat TextGrid files
---
## Dataset Structure
```
README.md # Corpus documentation
sessions.csv # Metadata for sessions and parts
speakers.csv # Metadata for speakers
Sessions/
G001/
G001_1_L.wav # Audio (left channel)
G001_1_L.TextGrid # Annotation (left channel)
G001_1_R.wav # Audio (right channel)
G001_1_R.TextGrid # Annotation (right channel)
G001_2_L.wav
G001_2_L.TextGrid
G001_2_R.wav
G001_2_R.TextGrid
...
G003/
G004/
G005/
G006/
G007/
G008/
G009/
G010/
G011/
G012/
G013/
```
- `sessions.csv`: Contains information about each session and part.
- `speakers.csv`: Contains information about speakers for each session, part, and channel.
### Naming Convention
Files under each session directory follow the naming scheme:
```
<session_id>_<part>_<channel>.{wav, TextGrid}
```
- `<session_id>`: Session identifier (e.g., `G001`)
- `<part>`: Part number within the session
- `<channel>`: Channel ID (`L` = left, `R` = right)
- Extension: `.wav` for audio files, `.TextGrid` for annotations
---
## Citation
If you use this dataset, please cite:
```bibtex
@Article{Mori2025,
author = {Mori, Hiroki and Kikuchi, Yuki and Okubo, Minato and Matsuda, Takuto and Shiratori, Keita and Arimoto, Yoshiko},
title = {Social Screams and Shouts in Action Gameplay Voice Chat: Annotation and Evaluation},
journal = {Language Resources and Evaluation},
year = {2025},
publisher = {Springer Nature},
note = {Under review},
}
```
---
## License
- **License**: [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/)
- Non-commercial use only.
- For commercial licensing inquiries, please contact the dataset maintainers.
---
## Contributing
For bug reports, feature requests, or suggestions regarding annotations,
please use the GitHub repository:
[https://github.com/hirokisince1998/AGSC](https://github.com/hirokisince1998/AGSC)
- **Issues**: For reporting problems or proposing new features/specifications.
- **Pull Requests**: For submitting corrections or improvements to annotations.
|