Reports suggest that Computer Vision is growing at a CAGR of 7.80% between 2018 and 2023. Tech giants are leveraging this technology in a number of ways, for instance, in autonomous driving and retail, among others.
Founded in 2015, Inkers.AI is one such startup which is playing a crucial role in solving vision and perception as a problem. IIT Alumni Rohan Shravan and Manish Giri are the masterminds behind Inkers’ designs and developments with deep large-scale AI and ML vision solutions for various sectors such as enterprises, retail, and defense & security organisations.
Analytics India Magazine caught up with Rohan Shravan, Founder and CEO of Inkers to gain insight on how the deep video processing startup is striving to leverage computer vision, deep neural networks, hardware-level acceleration, and special frequency domain encoders to create an extensive and integrated video processing pipeline for the users.
In the beginning, the company faced several challenges, mainly with vision while developing algorithms or other camera-based inputs due to the unavailability of graphic cards. Currently, the core tech can run the competition 10 to 20 times on the same hardware. The company has trained a lot of engineers in the last 2 years and we have trained more than 3,000 people for the programme called EIP
Inkers is a vision AI large-scale platform company for core Intelligent Video Analytics (IVA) applications like facial recognition, object detection, tracking, visual workflow automation, etc. Its platform, Ink vision is an AI operating system that provides AI outwork which means the base operating system provides face recognition, face tracking, object recognition object tracking and other data networks by default. Ink vision develops an application that helps to track customers and also develop the applications with the customers as well as provide them with the application.
Use Of AI/ML
Inkers is mainly focused on vision-based AI where most of the algorithms are based on deep networks. According to Shravan, the algorithm can perform at least 30% to 40% better than the traditional computer vision networks. In the last 6 months, the product has gone live and has around 4,500 clients at present and around 4,000 cameras deployed around Indian space.
The company has teamed up with several domains such as travel industry to provide them the system knowledge-based engine where the engineers can easily track problems that can be solved by the Aircraft construction building, Law Enforcement for major crime cases and help them with super-resolution of the faces of criminals as well as detecting a particular vehicle out of all the cameras across a city, Bangalore international airport for a tender for a lead management where the cameras are looking at all the lanes and giving information about the number of people in lane.
The company faces competition in the security and surveillance part. According to Shravan, there’s a company called Visionatics which is taking parts in many tenders in the field of computer vision. However, these companies aren’t making use of the current security networks which is why in terms of efficiency and performance Inkers generally beat them.
With the key focus on deployment scalability, and vision AI, the company aims to become the default Visual Perception Platform for other AI application developers or ISVs. Currently, the company is focusing on two industries, first, it is working on automation and next it is working on automating human resource management. In the next 5 years, the company is planning to become the first AI for vision computing.