Final Project Report: Project Report.pdf
- Trained two Convulational Neural Networks (CNNs) to recognize age and gender of a valid human facial images: one CNN for age detection, and one for gender detection.
- Dataset used: Adience Dataset - https://talhassner.github.io/home/projects/Adience/Adience-data.html
- Accuracies of Model: Age model - 43%, and Gender model - 80%
- Paper followed: Age and Gender Classification using Convolutional Neural Networks. https://talhassner.github.io/home/projects/cnn_agegender/CVPR2015_CNN_AgeGenderEstimation.pdf