TalkTriUnity is aimed at breaking down the barriers for disabled people to access virtual assistants seamlessly. We used GPT-4-Turbo, Whisper, Clarifai's Text to Speech models for our submission. Whisper for Speech to text and GPT4 Turbo to generate answers. Currently it supports only letter by letter sign language translation but our aim to support gesture based translation as well. Further, in future we also want to have the facility of Text to Sign Language translation. For the sign language translation we utilized the open source code available at https://github.com/emnikhil/Sign-Language-To-Text-Conversion/tree/main. The interface is user friendly enabling virtual assistants to be more inclusive and accessible.
Category tags:Lokesh Badisa
Team member not visible
This profile isn't complete, so fewer people can see it.
Ishita Bardhan
"Although I couldn't access the app, I liked the innovation in the project. Also, the presentation lacks business details and need a lot of improvement. Good luck! "
Muhammad Inaamullah
Machine Learning Engineer
"Excellent idea. It's very helpful however, I'm curious about the business model along more about the technology and flow which wasn't in the presentation. Good use of technology."
Muhammad Mubashir Hassan
Developer
"Unfortunately, couldn't access and test the app live. However, there is excellent use of many different models from Clarifai, and a short video effectively showcases the app's features. I wish there were more information on the business value, and the presentation could be much more detailed. But, in general, good work!"
Sumanth Papareddy
ML Developer Advocate