GeSpe system stands for Gesture Recognition and speech recognition system .
In the past few months, my Grandpas eyesight has been so poor he has been struggling to turn on the tv, turn it off, change the channel because he can’t see the remote clearly. So if I made a system he can be using gestures or voice to control it will mean a lot to him.
I have collected my training dataset from 20BN.jester. This dataset contains 200k+ images with 27 classes of gestures of which 5 classes of the gestures do nothing and the rest are functional. The voice recognition is to control all tasks and is able to recognize the old man’s voice.
I wanted to make the voice recognition only recognize Grandpa’s voice unless I added another voice but I don’t know how I think you guys can help. I am not yet done with the project as I also want to use it as my final project on campus, the moment I will be through I will share my files with you guys.
Mean while am open for any other idea. I will share my files soon am through.