AMLS1_Final

This Repository is used for assessment of AMLS1 This project requires external libraries: os, numpy, keras, cv2, dlib, matplotlib.pyplot, pandas and sklearn All essential functions are included in functions.py from data collection, feature extraction, hyper-parameter tuning to model. Main.py includes models for each task, by executing main.py, 4 test accuracies will be generated. However, if you wish to see how hyper-parameters are tuned, then you need to run Hyper_Paramter_Tune function separately and I don't suggest you to do it as it takes long time to run cross-validation. Importantly!!! This code will takes 10 mins for task A and probably 40 mins for task B as it contains larger dataset and complex model. !!! Importantly A1,A2,B1,B2 and Dataset are all empty, I just followed the structure you given in the assessment notes. Useful files are functions.py, main.py and shape.dat