Under the open-innovation theme, my team designed a novel approach to convert 2D facial images to 3D images by implementing Position Map Regression Networks(PRnet) and further used the concept to create animated faces in the wild for character animation in real time RPG Games. I implemented the PRnet and worked on developing a OpenGL based framework for Android to project the 3D faces.
Automatic creation of Interactive 3D Unity Assets of each segmented object(ex- Human face, human body, furniture, TV, Car, pets, etc) from a live camera clicked 2D image using Deep Learning.
Measuring Eye Lens power with a smartphone and eradicating the use of lensometer machines and external dependencies while using a mechanism similar to the machine on a smartphone utilizing Deep Learning principles. Our team worked together on developing an equation to calculate and measure the power of eye lens after in-depth and rigorous research in Optics and biology to utilize the focal data received from the models to calculate the power of the eye lens.
Using two 360-degree GoPros, FitBit and Deep Learning, our team developed a novel solution to detect enemy attack for soldiers in the field and detect incoming enemies and alert the control and the soldier. Presented to the Royal Canadian Navy.
I developed a novel solution, for which I currently hold the patent. I developed a method to detect, quantify, track and diagnose facial paralysis deseases such as Bell's Palsy, Ramsay Hunt's Syndrome and others. Received media coverage in Business Standard, Economic Times and others.
Developed an app for old age people for universal accessibility and on-demand help among other features(which I honestly not able to recall :p).
Our team developed an Internet of Things(IoT) enabled online and offline smart grid management system for the Government of India.