---
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: ipa
    dtype: string
  - name: text
    dtype: string
  - name: speaker_code
    dtype: string
  - name: speaker_gender
    dtype: string
  - name: speaker_age
    dtype: int64
  - name: pronunciation_accuracy_0_to_10
    dtype: int64
  - name: pronunciation_completeness_fraction
    dtype: float64
  - name: pauseless_flow_0_to_10
    dtype: int64
  - name: cadence_and_intonation_0_to_10
    dtype: int64
  splits:
  - name: train
    num_bytes: 137594253
    num_examples: 1120
  - name: test
    num_bytes: 130964304.625
    num_examples: 1123
  download_size: 253937800
  dataset_size: 268558557.625
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
license: apache-2.0
task_categories:
- automatic-speech-recognition
language:
- en
tags:
- Speech
- IPA
- Mandarin
pretty_name: speechocean762 filtered
size_categories:
- 1K<n<10K
---

# speechocean762
[speechocean762](https://www.openslr.org/101) is a speech dataset of native Mandarin speakers (50% adult, 50% children) speaking English.
It contains phonemic annotations using the sounds supported by [ARPABet](https://en.wikipedia.org/wiki/ARPABET).
It was developed by Junbo Zhang et al. Read more on their [official github](https://github.com/jimbozhang/speechocean762), [Hugging Face dataset](https://huggingface.co/datasets/mispeech/speechocean762), and [paper](https://arxiv.org/abs/2104.01378).

## This Processed Version
We have processed the dataset into an easily consumable [Hugging Face dataset](https://huggingface.co/docs/datasets/en/index) using [this data processing script](https://github.com/KoelLabs/ML/blob/main/scripts/data_loaders/SpeechOcean.py).
This maps the phoneme annotations to [IPA](https://en.wikipedia.org/wiki/International_Phonetic_Alphabet) as supported by libraries like [ipapy](https://pypi.org/project/ipapy/0.0.1.0/) and [panphon](https://pypi.org/project/panphon/0.5/).
We filter out samples with uncertain sound labels, unknown labels, and heavy accents for use in Speech-to-IPA modeling as opposed to pronunciation scoring which the original dataset should be used for.

**Note**: we have [another version of this dataset](https://huggingface.co/datasets/KoelLabs/SpeechOceanNoTH). 
The difference is that in this filtered version, we have removed samples where a different set of 2 annotators disagreed with the labels in the original.

- The train set has 1120 samples (around 72 minutes of speech).
- The test set has 1123 samples (around 68 minutes of speech).

All audio has been converted to float32 in the -1 to 1 range at 16 kHz sampling rate.

## Usage
0. Request access to [this dataset](https://huggingface.co/datasets/KoelLabs/SpeechOceanNoTH) on the Hugging Face website. You will be automatically approved upon accepting the terms.
1. `pip install datasets`
2. [Login to Hugging Face](https://huggingface.co/docs/huggingface_hub/en/guides/cli#huggingface-cli-login) using `huggingface-cli login` with a token that has gated read access.
3. Use the dataset in your scripts:
```python
from datasets import load_dataset

dataset = load_dataset("KoelLabs/SpeechOceanNoTH")
train_ds = dataset['train']
test_ds = dataset['test']

sample = train_ds[0]
print(sample)
```

## License
The original dataset is released under the Apache 2.0 license, a summary of the license can be found [here](https://www.tldrlegal.com/license/apache-license-2-0-apache-2-0), and the full license can be found [here](https://www.apache.org/licenses/LICENSE-2.0). 
This processed dataset follows the same license.