Facial Attribute Detection, Cellebrite

Sheryl Sitruk

Data Science Fellows June 2020 Cohort

Abstract

Detecting facial attribute and accessories on images

Challenges (at least two)

1. Working with a team over zoom
2. Dealing with computer vision problem 

Achievements (according to KPIs)

1. Around 95% accuracy with deep learning models
2. Around 90% accuracy using embeddings and classic machine learning models

Future project development 

During this project I managed a computer vision project from the beginning to the end. Starting by collecting, data, training and finally creating an API. 

This project will be used by Cellebrite for their internal use.