Inspiration
While watching a movie , i saw a few people trying to communicate with people who have difficulties in speaking , so they used hand gestures to communicate with them . And also a few more places where humans do interact with animals using hand gestures , got me drive into making this project.
What it does
This project uses your live camera to translate each ASL to english , so that people can communicate efficiently.
How we built it
Using python and a machine learning model , which work together to give output
Challenges we ran into
There are a various number of challenges such as limited amount of dataset , also we had to make our own custom dataset , processor to run the model which took almost 2 hrs to run and sufficient RAM consumption. Every library is not compatible with other , so needed to find the ones which work with other without any problem.
Accomplishments that we're proud of
We are Proud to make such a interface which does not divide people who can speak and can't. It united people and does not make non-verbal different from others.
What we learned
The most fun part on making this project was to learn the ASL , every language has it's own language and also how interesting it is to communicate with others
What's next for Sign Language Translation
There's more to as we can convert speech to text and also text to speech for people who don't know ASL to communicate with others . I have started building that also , Using API to fetch data to convert and other .
Built With
- jupyter-notebook
- keras
- lstm
- machine-learning
- opencv
- python
- tensorflow
Log in or sign up for Devpost to join the conversation.