Data-Science-Community-SRM/Hand-Gesture-Recognition-Rock-Paper-Scissor
Hand Gesture Recognition and Modification was based on transfer learning Inception v3 model using Keras with Tensorflow backend trained on 4 classes - rock, paper, scissors, and nothing hand signs. The final trained model resulted in an accuracy of 97.05%. The model was deployed using Streamlit on Heroku Paas.
Jupyter NotebookMIT