Why an ASL Translator?
Ever tried conversing overseas where English is not the native language? Well perhaps most people have a basic understanding of a little English, now imagine not being able to be understood at all, well that is the unfortunate circumstance for some.
Over 500,000 people in Singapore are hearing impaired or deaf, unfortunately, some can only converse in sign language. With almost every Singaporean having not learned sign language, it is almost impossible for those without the ability to speak to converse, attend job interviews, and hold presentations.
How does the ASL Translator work?
Through image and pattern recognition as well as machine learning, our program processes hundreds of frames in live time, searching through our database to then translate the sign , and finally displaying the corresponding word in English on the same screen, all in milliseconds.
Complications faced
There are over 10,000 unique signs, with most being action-based. The perfect system would include the database of all the signs and their corresponding words, however that would require the computer to search through hundreds of TB of data. The regular computer would require hours to translate one single sign. Therefore we have truncated the program to only a few signs to match our computing power.
The Future.
We hope to be able to upscale our ASL Translator to include every sign as well as the ability to translate into languages other than English. The ASL Translator ultimately aims to enable our beneficiaries to express themselves easily and freely as should be the right of every single human being.
Built With
- machine-learning
- pattern-recognition
Log in or sign up for Devpost to join the conversation.