Projects

Neutrino

Currently working on a project codenamed Neutrino to enable Aruba Networks to enter into the Smart Offices space. This project uses a Deep Learning (MobileNetSSD v1) model to detect people (primarily) and objects. This data is then fed via Kafka into a Cassandra database. A (Flask based) UI is used to visualize statistics about room, building floor or area usage. An example of useful statistics that can be gleaned can be seen below. This project is designed to use standard surveillance cameras as well Raspberry Pi based cameras.

The project is currently in alpha stage, with few basic applications fully implemented. Work is ongoing to expand the feature list as well as transition project into AWS.

Facial Recognition

Experimented with Face Recognition on Raspberri Pis. This project used OpenCV to manipulate images and pass it on to a DNN module that used pre-trained models to detect and identify faces. A RPi 3 could usually reliably identify faces in about 2 to 3 seconds.

Other Projects

List of other projects Coming Soon!