metadata

license: cc-by-4.0
task_categories:
  - automatic-speech-recognition
language:
  - en
size_categories:
  - 1K<n<10K
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: duration
      dtype: float64
    - name: sentence
      dtype: string
    - name: uid
      dtype: string
    - name: group_id
      dtype: string
  splits:
    - name: train
      num_bytes: 2266558522
      num_examples: 8000
    - name: valid
      num_bytes: 260170178
      num_examples: 1000
    - name: test
      num_bytes: 283817142
      num_examples: 1000
  download_size: 2784217234
  dataset_size: 2810545842
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: valid
        path: data/valid-*
      - split: test
        path: data/test-*

🎬 ScreenTalk-XS: Sample Speech Dataset from Screen Content 🖥️

📢 What is ScreenTalk-XS?

ScreenTalk-XS is a high-quality transcribed speech dataset containing 10k speech samples from diverse screen content.
It is designed for automatic speech recognition (ASR), natural language processing (NLP), and conversational AI research.

✅ This dataset is freely available for research and educational use.
🔹 If you need a larger dataset with more diverse speech samples, check out:
👉 ScreenTalk (Full Dataset)

📜 Dataset Details

Feature	Description
Total Clips	10k transcribed speech samples
Languages	English
Format	`.wav` (audio) + `.csv` (transcriptions)
Use Case	ASR, Speech-to-Text, NLP, Conversational AI

🆓 This dataset is free to use for research purposes.

📊 train set duration: 18.55 hours
📊 valid set duration: 2.12 hours
📊 test swet duration: 2.32 hours

📂 Dataset Structure

ScreenTalk-XS contains the following fields:

Column	Description
`audio`	Audio info
`duration`	the duration of the audio chunk
`sentence`	Transcribed speech

📌 Example Entry

{
    "audio": "path/to/audio.wav",
    "sentence": "I will find you and I will train my model."
}