xiD is an advanced attendance system that leverages cutting-edge computer vision technology to streamline and automate the process of tracking attendance. By utilizing facial recognition and image processing algorithms, xiD accurately identifies and records the presence of individuals in various settings such as classrooms, offices, and events.

Oversaw the optimization and overhaul of the Facial Recognition Based Attendance System deployed at Xavor's offices

Optimized system performance; reducing facial detection time from 0.02s to 0.004 s, reducing facial recognition time from 0.30s to 0.18s, reducing pose estimation time from 0.03s to 0.000008s, reducing system start up time from 14 s to 3.86 s, increasing number of people supported from 150 to 1000+; by updating existing Dlib detection and recognition models with custom arc-face models.

Converted the desktop-based implementation of the system to a modular web app decoupling the front end and the back end; enabling integration support with 10+ different HR management systems; using Rest APIs and web sockets.

Machine Learning, Data Analisys, REST/RESTful APIs, AWS Glue, Flask, Python, Docker, Modular APIs, Azure, ETL, Pandas, Numpy, Git