Deep-Learning model to detect "Age of the Actors" and classify them into three categories namely-Old, MIddle and Young- using facial feature images.
I was able to achieve 0.9436 accuracy with single model on the test set and 4th position out of 7516 at the leaderboard.
This repository contains the notebooks that i created for competing in the "AGE DETECTION" competition along with the submission files.
All the files are in the submission folder itself.
I took the competition as 10 day challenge and dropped the competition midway when the deadline was extended by 96 days.
here is the link for the competition:
https://datahack.analyticsvidhya.com/contest/practice-problem-age-detection/
further improvements can be done by:
- increasing augmentations.
- applying Test Time Augmentation
- Ensembling of my single models.