Event Report: Dr. Ramesh Raskar of MIT Media Lab
(This is a live blog of the talk given by Dr. Ramesh Raskar, of MIT Media Lab in Pune. Since this is a live-blog, it will not be as well structured as a regular article, and might contain more-than-normal grammatical errors.)
EyeNetra is a very small, cheap device (that costs less than Rs. 100) that can be clipped on to a regular smartphone and which can be used to detect vision problems including detecting their lens prescription, astigmatism, and even cataract. Since it is so cheap, and portable, it can be used in villages all over the world. In India alone, about 6% of the people wear glasses, but it is estimated that about 40% of them should be wearing them. That’s 200 million people in India who don’t have eye glasses that are needed.
Why is this a big deal? Blurry vision means that a child cannot learn. Blurry vision means that there are certain jobs that a day labourer cannot do. So EyeNetra can have tremendous social impact.
In EyeNetra, the software on the smartphone displays a number of dots on the screen. The clip on device has a number of tiny lenses which are placed in such a way that if you have normal vision, the light rays from all the dots will actually convert on the retina of your eye and you’ll see a single dot. If your eye has a problem, then you’ll see multiple dots. Now the UI of the software asks the user to adjust things until the dots converge and the user sees only one dot. Based on what adjustments are needed, the software will be able to figure out what are the defects in the user’s eye (in terms of spherical and cylindrical corrections)
EyeNetra needs high resolution displays, but in recent years the resolutions of phones have really gone up, from 160DPI for samsung to 300+ for the iPhone 4G. User demand is driving industry to improve the resolutions of their phone. So, every time you use your phone to see video clips and take photographs, you are forcing the industry to increase their resolutions, and will indirectly end up helping people around the world get better vision through EyeNetra.
Netra prototypes are now in dozen+ countries.
The next device in this series is EyeMotia – for detecting cataracts. It is a similar clip-on device for a smartphone which uses similar techniques to determine whether you have cataract. The basic idea is similar – the software draws various patterns on the screen which pass through a specific area of the lens in your eye before reaching the retina to form a clear green dot. If you have normal vision, you will see a simple green dot going around in circles. If you have astigmatism, you will see the green dot going around in an oval path. If you have cataract, the green dot will disappear at certain times as it goes round. This is because at a certain location, when it has to pass through a cataract affected portion of your lens, the rays will get scattered and will not form a nice green dot on the retina.
The eye is the only part of your body where you can see blood vessels directly without having to cut you up. Similarly, if you know what to look for, you can look into the aqueous humour (the colorless liquid in the eye), you can make deductions about the blood sugar levels in your body. So, the eye is an amazing device, and you can use clever visual computing to do various interesting deductions about your body using simple devices and smartphones.
EyeNetra is setting up a team in India which will work with hospitals, government organizations, NGOs and other groups to take the EyeNetra device to rural India. They tried just giving the devices away to NGOs, but that did not work well – so the current thinking is that it needs to be run like a business using a focused team for success. So, EyeNetra is looking for people who will join the team. A COO, maybe a CTO, BizDev are needed. Anyone interested should contact Ramesh.
Challenge to People – the smart phone is an amazing device. There is lots and lots you could do with it. Think of various ways in which you can use it for purposes that it was not originally intended for. There is the camera, the display, accelerometer, GPS, internet, bluetooth, RF. You can do magic.
Think of this example of thinking out of the box: create a video game in which people with normal vision will shoot one way, and people with abnormal vision (astimatism, color-blindness) will shoot a different way. So you get a medical test done while playing a video game.
For more information about EyeNetra, see http://EyeNetra.com
Why Visual Computation Will be Big
- In the next few decades, the world will move from text and audio based communications to more and more visual information. Vision crosses language barriers, socio-economic barriers, and will help the next billion consumers. Hence, processing visual information intelligently becomes a very important capability.
- In 6 years, the world went from zero cameras in mobile phones, to a billion cameras in mobile phones. And today, a billion mobile phones with cameras get sold every year. There is a major visual revolution underway, but most people haven’t realized it yet.
- Hence, the Camera Culture group spends their time exploring various ideas related to visual computing. They spend 60% of their time on hardware and 40% on software. With this, they build crazy cameras – like the camera that can look around corners.
- Looking around corners: How is this done? Use the flash from a camera. The light hits a wall/door/obstacle and bounces off in various directions. Some of the bounced photons actually go around the corner, hit various objects that are not directly visible, and then an even smaller fraction of them bounce back all the way to the camera. If you’re clever about analyzing the photons, you can actually figure out where each photon has come from and hence reconstruct features of the objects around the camera. For this you need to do an extremely fast camera – which does one trillion frames per second.
- If you do the work that you’re supposed to be doing, and then spend a little more time doing ‘something extra’, that something extra has a high chance of being noticed. So everybody – do your job well, but make sure to do something extra
- In a way, it is good to work in an emerging country like India. Here, you are not totally constrained by draconian governmental regulations that limit your creativity and possibilities. Of course, we also have regulations, but they’re not as strong, and not as strongly enforced. Hence, you can achieve much more here, and more quickly than what would be possible in the US. In fact, you can help people more because the Government is staying out of the way.
- MIT has a $100k Entrepreneurship & Ideas competition every year. This has 3 stages. A 1-minute elevator pitch contest in October, with $1000 instant prizes, followed by a Executive Summary competition in November, with $1000 instant prizes, followed by a full-fledged Business Plan competition in Jan/Feb which has various track prizes, and a grand prize of $100k. Tip: get on their mailing list and you can get an idea of everything that’s going on. So that is something worth doing.
- If Pune would like to start such competitions Ramesh is willing to put in some money from his Entrepreneurship class (Imaging Ventures) to fund the competition.
- There are dozens and dozens of classes in MIT for converting innovation to commercial success. This includes basic+applied research all the way to classes targeting people in established companies. What you can do, sitting in Pune, is join the mailing lists of these classes, and see the course material on the web. For free.
- Thinking about difference between Pune and Boston (MIT) – the same people who don’t do much here go to Boston and do amazing things. What is the difference? Network. Everybody has to go out of their way to help other people in the network – and this has a huge multiplier effect.