Information for the task can be found here
Dataset here
Further pretraining of pretrained InceptionV3 (imagenet) using Celeb A dataset before finetuning on Adience. Refer to paper.pdf for detailed report.
Results:
- Gender: 93.55%
- Age: 55.04%
Note: some .ipynb are duplicated to run different folds of the data in parallel using Kaggle kernel.