HyperVerge, a deep-learning start-up which is based in Silicon Valley and India, has raised $1 million in a seed funding round from New Enterprise Associates (NEA), Milliways Ventures and Naya Ventures.
With the infusion of new capital, the company will now expand its operations to other cities in US as well as in India. Currently, it has offices in Palo Alto, Bangalore, and Chennai.
HyperVerge Inc. was founded on July 25th 2014 by a group of students which include Kedar Kulkarni, Vignesh Krishnakumar, Kishore Natrajan, Saivenkatesh Ashokkumar and Praveen Kumar.
They started working on prototype of their cloud-based image recognition technology while at IIT Madras’ Incubation Cell. They have advisors from MIT Media Labs and the IIT alumni community.
By using deep-learning and computer vision, it has developed a cutting-edge technology for recognition of people, scenes, events, documents, objects within photos and for analyzing photo quality to find bad photos, duplicates, spam photos, etc. They have filed a patent for this which is currently pending.
It uses deep learning algorithms to process consumer images and videos on the cloud. All these technology modules are coming together in an app called Silver. The upcoming mobile application will help users automatically organize all the images on phone.
Kedar Kulkarni, CEO and co-founder of HyperVerge, said,
Working with a budget of under 6 lakhs, we were able to create classifiers for identification of people, scenes, events and unique patterns in images. IT Madras’ Incubation Cell and Alumni community have provided tremendous support to our team. With our core team and investors in place, we believe that our image recognition technology can be the foundation for several breakthrough consumer applications in the near future.
Kittu Kolluri, General Partner at NEA, said,
The technology developed by the company can span out as useful applications for consumers, enterprises and developers. We are thrilled to partner with HyperVerge to bring innovative deep-learning solutions to users and prospective partners.
