CSC3095 - Evaluating deep learning approaches for perceived age prediction

Usage

python train_age_classification.py/train_age_regression.py --b 512 --lr 0.001 --layers 8 --type 'VGG16' --mode 'finetune'

Datasets

These can be placed in the data folder and processed with the relevant scripts in util/ and face_detection/

Mode

• pretrain - IMDb-Wiki dataset
• finetune - ChaLearn Percieved Age V2 dataset

TensorBoard logging

Can be switched on by adding the callbacks paramater to the model.fit() with the callbacks variable

References

• Rothe, R., Timofte, R. & Van Gool, L. (2016) Deep Expectation of Real and Apparent Age from a Single Image Without Facial Landmarks. International Journal of Computer Vision. 126 (2-4), 144–157.
• https://data.vision.ee.ethz.ch/cvl/rrothe/imdb-wiki/
• http://chalearnlap.cvc.uab.es/dataset/19/description/
• https://github.com/ageitgey/face_recognition_models