Japanese STS Translation Arena: Speech-to-Text Model Comparison
Compare different speech-to-text models for Japanese and vote for the best one. Hear the translation with Kotoba TTS.
This arena is inspired and built on TTS Arena.
We are actively maintaining this project. Suggestions via contact/discussion are welcome!
🎤 Speech-to-Text Model Comparison
- Input Japanese speech by uploading an audio file or recording with your microphone.
- Select a target translation language.
- The system will process your speech with two different speech-to-text models.
- Compare the results side by side and vote for the better translation.
- Listen to both translations spoken by Kotoba TTS.
Note: It may take up to 30 seconds to process your speech and synthesize audio.
🏆 Leaderboard
Vote to help the community determine the best Japanese speech-to-text (STT) models.
The leaderboard displays models in descending order of translation quality (based on votes cast by the community).
Important: In order to help keep results fair, the leaderboard hides results by default until the number of votes passes a threshold. Tick the Reveal preliminary results to show models without sufficient votes. Please note that preliminary results may be inaccurate.
Show all models, including models with very few human ratings.
📄 About
The Japanese STS Translation Arena allows users to compare different speech-to-text models for Japanese. It is inspired by TTS Arena.
Features
This application allows users to:
- Input Japanese speech through microphone recording or file upload
- Compare two different speech recognition models side by side
- Select a target language for translation
- Vote for the better translation
- Listen to the translated text spoken by Kotoba TTS
Credits
Thank you to the open-source code from TTS Arena which helped make this project possible.
Request a model
Please create a Discussion to request a model.
Privacy statement
We may store speech you input, recognized text, and generated audio. We store a unique ID for each session. You agree that we may collect, share, and/or publish any data you input for research and/or commercial purposes.
License
Generated audio clips cannot be redistributed and may be used for personal, non-commercial use only.
If you use this data in your publication, please cite us!
Copy the BibTeX citation to cite this source:
@misc{sts-arena-ja,
title = {Japanese Speech to Speech Arena},
author = {Kotoba Technologies.},
year = 2024,
publisher = {Hugging Face},
howpublished = "\url{https://huggingface.co/spaces/kotoba-tech/STS-Translation-Arena}"
}
Please remember that all generated audio clips should be assumed unsuitable for redistribution or commercial use.