Learning language systems are expensive or don’t provide immediate feedback to learners. Zah Hah Is a open language learning system, the goal is to leverage modern speech recognition engines to assist non-native speakers in learning a new foreign language or native speakers with speech impediments. The initial prototype will use Siri’s (iPhone) speech recognition engine to process users responses to prompts initiated by a learner and message this back using the Twilio SMS service. The goal is to make learning languages free and to provide valuable analytical data from the learner’s interactions with the system. The system will prescribe words and phrases that were previously identified as problems, we will determine these based on longer response times which should be an indication that the speech recognition engine (Siri) didn’t recognize what the user was attempting to say.
Problem is concerned with
Find the source and information on https://github.com/boryaku/Zah-Hah
In schools at home for practice.
User initiates a session with Zah Hah the system responds back with words and phrases that are a part of the program. The user then accurately replies back to the system with the supplied word/phrase.
The user views his reports on trouble phonemes and overall progress
The system prescribes additional words or phrases based on trouble items.
Open Speech Recognition engines with a large vocabulary (siri works pretty good and will get better), SMS costs or messaging costs add up as user join the system.
Make it work from a browser too, extend the speech recognition capabilities to the browser.
Instructional Design, speech recognition systems, multi language speakers, teachers, developers, speech therapists.
Not yet..
Add new comment