A hand gesture & Sign Language Detection Project, based on Computer Vision. Action Recognition using MediaPipe Holistic, OpenCV and an LSTM Deep Learning model.
Here, this model is trained manually for three actions:
- "Hello"
- "Thanks"
- "I Love You"
Trained on cpu, here are the following insights from tensorboard while training
- Test data : 1
- Training data : 0.9294117647058824