Audiobox Aesthetics

🚀 Audiobox Aesthetics

Unified automatic quality assessment for speech, music, and sound.

This model has been pushed to the Hub using the PytorchModelHubMixin integration:

• Code: https://github.com/facebookresearch/audiobox-aesthetics
• Paper: https://huggingface.co/papers/2502.05139

--- README below copied from https://github.com/facebookresearch/audiobox-aesthetics

🚀 Quick Start

Audiobox Aesthetics offers unified automatic quality assessment for speech, music, and sound. You can quickly start using it by following the installation and usage steps below.

✨ Features

• Unified quality assessment for speech, music, and sound.
• Multiple pre - trained models available.
• Support for prediction via CLI and Python script.
• Released evaluation dataset with aesthetic annotation scores.

📦 Installation

1. Install via pip

pip install audiobox_aesthetics

2. Install directly from source

This repository requires Python 3.9 and Pytorch 2.2 or greater. To install, you can clone this repo and run:

pip install -e .

📚 Documentation

Pre - trained Models

💻 Usage Examples

Basic Usage

How to run prediction using CLI:

1. Create a jsonl files with the following format

{"path":"/path/to/a.wav"}
{"path":"/path/to/b.flac"}
...
{"path":"/path/to/z.wav"}

or if you only want to predict aesthetic scores from certain timestamp

{"path":"/path/to/a.wav", "start_time":0, "end_time": 5}
{"path":"/path/to/b.flac", "start_time":3, "end_time": 10}

and save it as input.jsonl

2. Run following command

audio-aes input.jsonl --batch-size 100 > output.jsonl

If you haven't downloaded the checkpoint, the script will try to download it automatically. Otherwise, you can provide the path by --ckpt /path/to/checkpoint.pt

If you have SLURM, run the following command

audio-aes input.jsonl --batch-size 100 --remote --array 5 --job-dir $HOME/slurm_logs/ --chunk 1000 > output.jsonl

Please adjust CPU & GPU settings using --slurm-gpu, --slurm-cpu depending on your nodes.

3. Output file will contain the same number of rows as input.jsonl. Each row contains 4 axes of prediction with a JSON - formatted dictionary. Check the following table for more info:

Output line example:

{"CE": 5.146, "CU": 5.779, "PC": 2.148, "PQ": 7.220}

4. (Extra) If you want to extract only one axis (i.e. CE), post - process the output file with the following command using jq utility:

jq '.CE' output.jsonl > output-aes_ce.txt

How to run prediction from Python script or interpreter

1. Infer from file path

from audiobox_aesthetics.infer import initialize_predictor
predictor = initialize_predictor()
predictor.forward([{"path":"/path/to/a.wav"}, {"path":"/path/to/b.flac"}])

2. Infer from torch tensor

from audiobox_aesthetics.infer import initialize_predictor
predictor = initialize_predictor()
wav, sr = torchaudio.load("/path/to/a.wav")
predictor.forward([{"path":wav, "sample_rate": sr}])

Evaluation dataset

We released our evaluation dataset consisting of 4 axes of aesthetic annotation scores.

Here, we show an example of how to read and re - map each annotation to the actual audio file.

{
 "data_path": "/your_path/LibriTTS/train-clean-100/1363/139304/1363_139304_000011_000000.wav", 
 "Production_Quality": [8.0, 8.0, 8.0, 8.0, 8.0, 9.0, 8.0, 5.0, 8.0, 8.0], 
 "Production_Complexity": [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0], 
 "Content_Enjoyment": [8.0, 6.0, 8.0, 5.0, 8.0, 8.0, 8.0, 6.0, 8.0, 6.0], 
 "Content_Usefulness": [8.0, 6.0, 8.0, 7.0, 8.0, 9.0, 8.0, 6.0, 10.0, 7.0]
}

1. Recognize the dataset name from data_path. In the example, it is LibriTTS.
2. Replace "/your_path/" into your downloaded LibriTTS directory.
3. Each axis contains 10 scores annotated by 10 different human annotators.

📄 License

The majority of audiobox - aesthetics is licensed under CC - BY 4.0, as found in the LICENSE file. However, portions of the project are available under separate license terms: https://github.com/microsoft/unilm is licensed under MIT license.

🔖 Citation

If you found this repository useful, please cite the following BibTeX entry.

@article{tjandra2025aes,
    title={Meta Audiobox Aesthetics: Unified Automatic Quality Assessment for Speech, Music, and Sound},
    author={Andros Tjandra and Yi - Chiao Wu and Baishan Guo and John Hoffman and Brian Ellis and Apoorv Vyas and Bowen Shi and Sanyuan Chen and Matt Le and Nick Zacharov and Carleigh Wood and Ann Lee and Wei - Ning Hsu},
    year={2025},
    url={https://arxiv.org/abs/2502.05139}
}

🙏 Acknowledgements

Part of the model code is copied from https://github.com/microsoft/unilm/tree/master/wavlm.