Inspiration
Many of us are familiar with the difficulty of learning a new language. There are many language education tools available, but short of taking a linguistics class, there are few options for learning proper pronunciation. This can be an integral component of really being able to effectively communicate in different regions. In some cases, even when aware of proper pronunciation, people can struggle to make sounds foreign to their native language. The rolling r's in Spanish are a common example. We wanted to create a language education app that has a larger focus in learning, and perfecting, pronunciation.
What it does
Phonemes is an English learning app that records user speech, and returns how well you pronounced each phoneme. A phoneme being the smallest unit of sound composing words. It keeps track of which specific phonemes a user struggles with the most and return words that will give the best opportunity to improve upon their weaknesses.
How we built it
Phonemes currently runs on a web app, built in HTML and CSS. The back end comprises of two parts: audio processing and problem word generation. The audio processing uses the SpeechAce API and Python to take in an audio file and target word to return scoring on the individual phonemes within the word. The word problem generation converts English words to IPA (International Phonetics Alphabet), produces a word of a specified difficulty that includes phonemes with which the user has trouble, and translates it back the the user's native language for prompting.
Challenges we ran into
The largest challenge we faced was with attempting to dynamically implement the SpeechAce API. We were able to access the API through a bash script, but struggled using any other language that would allow easier modification of input fields.
Accomplishments that we're proud of
We believe that we have produced a strong proof of concept for an app that solves an important problem in language acquisition.
What we learned
Before this event, no one on our team had Web Development experience. In addition to learning how to build a web app, we learned how to process audio files and interface with a new API.
What's next for Phonemes
1) Integrate capabilities for more languages 2) Expand database of words for question generation 3) Create a mobile app to make program more accessible 4) Create an open source database for community collaboration of word pronunciation examples
Built With
- css
- englishtoipa
- google-translate
- html
- javascript
- python
- speechace
Log in or sign up for Devpost to join the conversation.