README.md

metadata

ymi_version: 1
language:
  - en
tags:
  - speech
  - pronunciation
  - error-detection
  - forced-alignment
license: cc-by-nc-4.0
pretty_name: 'EPADB: English Pronunciation Assessment Dataset'
size_categories:
  - 1K<n<10K
task_categories:
  - automatic-speech-recognition
  - audio-classification
task_ids:
  - speech-recognition-other
  - audio-classification-other

EPADB: English Pronunciation Assessment Dataset for Batched Diagnosis

Dataset Summary

EPADB contains curated pronunciation assessment data collected from Spanish-speaking learners of English. Each utterance has manually aligned phone-level annotations from up to two expert annotators along with per-utterance global proficiency scores. Metadata links the aligned phones with MFA (Montreal Forced Aligner) timestamps, derived error classifications, and reference transcriptions. The corpus ships with a train and a test partition and includes speaker-wise waveform recordings resampled to 16 kHz.

Supported Tasks

• Pronunciation Assessment – predict utterance-level global scores or speaker-level proficiency tiers.
• Phone-level Error Detection – classify each phone as insertion, deletion, distortion, substitution, or correct.
• Alignment Analysis – leverage MFA timings to study forced alignment quality or to refine pronunciation models.

Languages

• L2 utterances: English
• Speaker L1: Spanish

Dataset Structure

Data Instances

Each JSON entry describes one utterance:

• Phone sequences for MFA reference (reference) and annotators (annot_1, optional annot_2).
• Phone-level labels (label_1, label_2) and derived error_type categories.
• MFA start/end timestamps per phone (start_mfa, end_mfa).
• Per-utterance global scores (global_1, global_2) and propagated speaker levels (level_1, level_2).
• Speaker metadata (speaker_id, gender).
• Audio metadata (duration, sample_rate, wav_path) plus the waveform itself.
• Reference sentence transcription (transcription).

Data Fields

Field	Type	Description
utt_id	string	Unique utterance identifier (e.g., spkr28_1).
speaker_id	string	Speaker identifier.
sentence_id	string	Reference sentence ID (matches reference_transcriptions.txt).
phone_ids	sequence[string]	Unique phone identifiers per utterance.
reference	sequence[string]	MFA reference phones.
annot_1	sequence[string]	Annotator 1 phones (- marks deletions).
annot_2	sequence[string]	Annotator 3 phones when available, empty otherwise.
label_1	sequence[string]	Annotator 1 phone labels ("1" correct, "0" incorrect).
label_2	sequence[string]	Annotator 3 phone labels when present.
error_type	sequence[string]	Derived categories: correct, insertion, deletion, distortion, substitution.
start_mfa	sequence[float]	Phone start times (seconds).
end_mfa	sequence[float]	Phone end times (seconds).
global_1	float or null	Annotator 1 utterance-level score (1–4).
global_2	float or null	Annotator 3 score when available.
level_1	string or null	Speaker-level proficiency tier from annotator 1 ("A"/"B").
level_2	string or null	Speaker tier from annotator 3.
gender	string or null	Speaker gender ("M"/"F").
duration	float	Utterance duration in seconds (after resampling to 16 kHz).
sample_rate	int	Sample rate in Hz (16,000).
wav_path	string	Waveform filename (<utt_id>.wav).
audio	Audio	Automatically loaded waveform (16 kHz).
transcription	string or null	Reference sentence text.

Data Splits

Split	# Examples
train	1,903
test	1,263

Notes

• When annotator 3 did not label an utterance, related fields (annot_2, label_2, global_2, level_2) are absent or set to null.
• Error types come from simple heuristics contrasting MFA reference phones with annotator 1 labels.
• Waveforms were resampled to 16 kHz using ffmpeg during manifest generation.

Data Processing

1. Forced alignments and annotations were merged to produce enriched CSV files per speaker/partition.
2. create_db.py aggregates these into JSON manifests, adds error types, and resamples audio.
3. Global scores are averaged per speaker to derive level_* tiers (A if mean ≥ 3, B otherwise).

Licensing

• Audio and annotations: CC BY-NC 4.0 (non-commercial use allowed with attribution).
• Please ensure any downstream usage complies with participant consent and institutional policies.

Citation

@article{vidal2019epadb,
  title   = {EpaDB: a database for development of pronunciation assessment systems},
  author  = {Vidal, Jazmin and Ferrer, Luciana and Brambilla, Leonardo},
  journal = {Proc. Interspeech},
  pages   = {589--593},
  year    = {2019}
}

Usage

Install dependencies and load the dataset:

from datasets import load_dataset

# Local usage before uploading:
ds = load_dataset(
    "epadb_dataset/epadb.py",
    data_dir="/path/to/epadb",  # folder with train.json, test.json, WAV/
    split="train",
)
print(ds)
print(ds[0]["utt_id"], ds[0]["audio"]["sampling_rate"])  # 16000

# After pushing to the Hugging Face Hub:
# ds = load_dataset("JazminVidal/epadb", split="train")

Acknowledgements

We thank the learners and expert annotators who contributed to EPADB, as well as the speech processing community for tools such as MFA and ffmpeg used in the data preparation pipeline.