simple opencv based project for HackNTU 2019 A simple software that has the power of 4
- Face detection and Identification 2)Training of image data 3)Emotion classification
- Audio classification and prediction
It provides as an interface and can be used as a security tool, an entertainment tool, an acting practice tool etc. It is interfaced to firebase to indulge in the internet of things.
open login.py and username = Rahul, password = Password
train = trains data of images into yml files to predict faces predict = predicts emotions and classifies face take data = crops faces and creates a directory of face images Audio = Can parse audio and get text of audio in most languages. Can also count the number of occurences of words
#We trained our own model but it was not accurate enough (as we could not train it long enough due to hackathon constraints) and we used a pretrained model and accompanying code to implement the model from a github repo that I do not remember the link of (we were very stressed and last minute). Though I will add it soon.