metadata

dataset_info:
  features:
    - name: system
      dtype: string
    - name: file_id
      dtype: int64
    - name: audio
      dtype: audio
    - name: text
      dtype: string
    - name: sound_quality
      dtype: int64
    - name: silence
      dtype: int64
    - name: pronunciation
      dtype: int64
    - name: worker_id
      dtype: int64
  splits:
    - name: train
      num_bytes: 1370581078.4
      num_examples: 9600
  download_size: 141788724
  dataset_size: 1370581078.4
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
language:
  - th
size_categories:
  - 1K<n<10K

ThaiMOS (TTS MOS Evaluaution)

(Older) TTS synthesized speech with human evaluation
Mean Opinion Score (MOS)
Annotation was done by Datawow
Annonation aspect: sound quality, pronunciation, silence
This dataset was originally developed in 2024 based on older TTS models -- likely that patterns in this data may not be applicable to modern TTS systems.

Annotation Guideline

In directory pack, there are 12 directories each with 50 utterances.

Each subject carefully listens to an utterance and give the scores in three aspects as follows.

Sound_quality (Noise level): คุณภาพของไฟล์เสียงว่ามีเสียงรบกวนหรือเสียง noise ต่างๆมากน้อยขนาดไหน
Silence: การเว้นจังหวะหายใจระหว่างประโยคและระหว่่างคำ
Pronunciation: การออกเสียงในแต่ละคำว่าออกเสียงได้ถูกต้องในระดับไหน

All aspects are assessed from 1 to 5 (The higher, the better).

Number of human subjects used for listening each utterance: 5-8 subjects
Evaluation Conditions
1. The subjects have to be born and raised in Bangkok.
2. The subjects have to be in quiet place to evaluate the speech audio files.