• 

  Our main screen

HackMIT2019

For Speech to Text, in the python_examples folder, run: pip install -r requirements.txt

Backstory

For our project, we were inspired by one of our team member's difficulty in learning Spanish. Reading a language you don't understand can be difficult. Mispronunciation is a common side effect of reading words you don't understand - how can you remember something if you don't know what it means. For our project, we wanted to make a tool for people to input sentences they want to learn to read aloud and then practice them with the computer's assistance.

Structure of the Tool

In order to use our tool, the learner must first input any sentence they want to practice reading. Then they press "start speaking" and read back their sentence to the computer. The computer will then highlight the words they got wrong and provide an image and the correct pronunciation of the word they got wrong. This way, the combination of audio, visual, and semantics will aid in a true understanding of they are saying.

Main Tools Used

Our webapp is a combination of HTML, Flask, and python code using several different APIs. For the speech-to-text portion of our app, we utilized REV.ai's streaming API. For the image search of incorrect words, we used Bing's image search API. Finally, to get the correct pronunciation, we used IBM's text-to-speech API.

Built With

• html
• python