A machine learning project related to object detection, facial emotions recognition and age/gender estimation. Currently the project contains files for training 2 models : Emotion model and Age-gender model; fully made demo is situated in
ahegao_trackerfolder with latest netowrks; code for the distillation process; script to crop faces for making dataset and finally the converted models that are ready to be transfered to mobile devices (converted tflite dir).
NOTE: The application was presented at AI BOOSTER 2020 conference. For video of the conference with our performance, refer to this link: https://youtu.be/CykzWldi9go (please note, that performance was done in Ukrainian and Russian)