Click the picture to get the video demo.
This is a project that allows users to record their speech, automatically convert it to text, and then translate the text to the given language. I created this project because coming from a family that doesn't speak english, my family often doesn't communicate well with people speaking English. I created this app in hope to assist their transition to using English to communicate. Although this app is definitely not perfect, I learned a lot of libraries related to speech-to-text. Also, I utilized the Kivy framework to create an app, which is really fun in the process too.
To run the application:
python -m app
Then press the button called record then start speaking.
If you don't want to run the application with a GUI, use:
python other.py
Then start speaking. Press the keyboard letter 'r' when you are done speaking.
- Google Trans
- Keyboard
- Kivy
- Numpy
- Scipy
- Sound Device
- Sound File
- SpeechRecognition
After you clone the respository, navigate to the project directory and run:
pip3 install -r requirements.txt
- Add a timer so the user know when the recording will end
- Create a better UI
- Allow custom duration of recording instead of constant recording
For any bugs and advices, contact me at [email protected].
Special thanks to Terry Kim for debugging some bugs for this project
Released under the MIT License.
The STHeiti Medium.ttc.ttf file comes from this repository.