Hi Reddit!

Very excited to do this AMA with you all. PROOF: http://imgur.com/s9sZBHx

You may recognize me from my TED Talk (Trillion Frames Per Second Imaging), our innovations in India via Kumbha Mela and REDX, my work in computational imaging and photography, or our eye prescription company, EyeNetra. Could not have pulled any of this off without the support of my research group: Camera Culture at MIT Media Lab.

Also, 3 ways to get involved with our work at MIT:

  1. Open positions in my group at the MIT Media Lab! Apply Here

  2. Jan 22-28th, 2016. If you’re in India, apply for our upcoming REDX Innovation Workshop at the end of January! We’re looking innovators from all walks of life to help us solve important problems in India. Also apply for Kumbhathon (http://kumbha.org). Applications close soon.

  3. Our company, EyeNetra, is providing solutions for better eye care. For this AMA, we’re offering 20% off the NETRA Autorefractor + Phone anytime today using code AMA_RAMESH.

AMA!

Comments: 328 • Responses: 48  • Date: 

Minifig81170 karma

I love the video of you guys shooting light into a Coke bottle.

I'm kind of curious; aside from the Tomato (or was it an apple) and the Coke bottle, what is the coolest thing you guys discovered doing this, that you guys didn't expect at all?

rraskar112 karma

Very good question. Even today, when I see the coke bottle video (http://web.media.mit.edu/~raskar//trillionfps/) I see new intriguing details. The video was captured by our scientist Andreas Velten and our large team.

There is a reversal of the sequence of events recorded in the video compared to the real world!

You can see great details in this publication below. This was not covered in the popular media. Andreas Velten, Di Wu, Adrian Jarabo, Belen Masia, Christopher Barsi, Everett Lawson, Chinmaya Joshi, Diego Gutierrez, Moungi G. Bawendi, and Ramesh Raskar. Capturing and Visualizing Light in Motion. SIGGRAPH 2013

PDF link: http://giga.cps.unizar.es/~diegog/ficheros/pdf_papers/femto.pdf

rraskar49 karma

You can see many more videos you may not have seen in popular media here: http://www.mit.edu/~velten/press/content/

DROP_DATABASE_USER78 karma

If there was one law of physics that you could "tweak," what would you change and how?

rraskar100 karma

Negative light!

adityassripada55 karma

Hello Prof. Raskar, I have attended Kumbhathon 3 and 4

I have 2 questions 1) I would like to know will there be a technology to make the blind see, like the bionic eye prototypes, in the market anytime soon ? Like in the next 4-5 years ? 2) Do you think it's possible to create a machine-brain interface which can feed data straight into human brain in non-invasive ways ?

rraskar81 karma

You should join our efforts to address preventable blindness at http://lvpmitra.com in collaboration with LVP Eye Institute in Hyderabad.

It is very much possible to create a bionic eye. But my guess is that bionic eye will be more effective for augmented reality than for any clinical applications. There are other shortcuts you will see. (1) Prevention: diagnostics and procedures to prevent blindness. This is probably the most exciting and impactful area. (2) Using other parts of retina by clever optics (3) Sensory substitution: Map visual signals to other human senses like touch or sound. (4) SynthBiology: channelrhodopsin manipulation on the retinal pigment to make it light sensitive (5) Neural stimulation bypassing the optical path and more.

AbAr201531 karma

Hi Dr. Raskar. Is it possible for a working professional like me to contribute towards the cutting edge research ongoing at the Media Lab remotely over weekends? Does Media Lab has any open code repositories where I could start from? I have a deep interest in Computer Vision and Deep Learning and I don't want that to trickle away due to my Mon-Fri job. Also, do we have any periodic G-Hangouts where I could connect?

rraskar33 karma

Come to our monthly Imaging Cafe and Health nights

http://www.meetup.com/Imaging-Cafe-Researchers-Entrep-Investors-MIT-Media-Lab/

http://www.meetup.com/ReDX-The-Boston-Health-Technology-Meetup/

We are hosting one in Bay Area in early Spring, so register on these websites for now.

rraskar16 karma

You can join our newsletter here and will announce many opportunities https://www.facebook.com/cameraculture?_rdr=p

tejaswikasarla20 karma

Hi Prof. Raskar, I've been an attendee at ReDX this July at Hyderabad and subsequently joined the innovation center full-time. I have watched your TED talk on EyeNetra a long ago. There you've described Netra as a 1$ smartphone clip-on eye piece for testing which I thought was a revolutionary idea in low cost health care. But now, Netra is a 2999$ kit. Can you please explain why this huge difference in terms of cost happened? More importantly, why is Netra not a 1$ smartphone clip on anymore? Though it still is a low cost screening device especially for rural areas, what factors affected the price?

rraskar26 karma

Very good question. The cost per test is still well below $1. The device sold at http://store.eyenetra.com solves many more problems and geared towards those specific professionals. It gives you not only refractive error but many other parameters. Plus it provides end to end HIPPA complaint online software solution. The smartphone itself is included in the cost. So the final cost includes solutions for those professionals. Btw you can get for $800 the full Netra-G including smartphone on the website today.

With scale, the cost will decrease over time beyond what you see for the first 100K units.

Oyyeee15 karma

Hi Professor,

What degree do you see being in the most demand over the next 5-10 years?

rraskar28 karma

Anything that combines the 3: life sciences (bio, biomedical engineering, bioinformatics, etc) data driven techniques (statistics, machine learning etc) and Emerging Worlds. Knowledge from degrees that were sufficient on their own (computer science, EE, medicine, MBA etc) will become necessary (you need to acquire that knowledge in your own time) but not sufficient.

gu1d3b0t14 karma

Have you done anything with light fields?

rraskar22 karma

Yes we have used lightfields not only for photography but for measuring eyeglasses prescription (http://eyenetra.com), for 3D displays (http://web.media.mit.edu/~gordonw/courses/ComputationalDisplays/), CAT scan machines (http://raskar.info/cat), 3D tracking (http://raskar.info/prakash) and more

lord_iceman8 karma

Which is your favorite Bollywood movie ?

rraskar17 karma

takenthatusernameis6 karma

Hi Prof. Raskar,

I am a recent Computer Science graduate and have a keen interest in Computer Vision and Computational Photography. I attended the MIT ReDX workshop held in Mumbai this year (Achuta was my mentor) and it was a great learning experience. I really appreciate MIT Media Lab setting up innovation centers in India. They actually get something done!

I'd really like to pursue my graduate studies in this field and it goes without saying that Camera Culture group is the perfect place to be at. However, I couldn't apply this year because of my ordinary profile. Here are some questions from my side:

a) How can a person like me (average grades, not much research experience but highly relevant work experience) improve his/her profile and chances to get into your group? I apologize if you'd be answering this question for the zillion-th time.

b) How much weight do grades carry when you are shortlisting candidates? Because IMO, grades surely do not reflect your potential and quality because the education system here is severely flawed, even in top colleges. (I'm from a Tier 1 college!)

c) Do you recommend any other schools/labs that are working on really cool camera related projects?

d) Can a person work in Camera Culture group in a non-student capacity? (As an assistant/associate etc.)

Thanks and keep up the awesome work you guys are doing!

rraskar23 karma

G-R-O-W .. I often get emails from highly motivated young men/women who want to pursue a dream but are asking for guidance on how to get started/where to find good ideas/which skills to learn and how to make the connections .. Here is what I often write back .. But send me your thoughts on how YOU would guide a young man or woman ..

Go breadth first: learn what is out there, from online classes, watch videos, read tech magazines

Rebuild you skill set: build stuff, write programs, do experiments, join local tech hobby grps, meet like minded people, help others build stuff in online communities

Online reputation: create a portfolio, make a website, write a blog about your thoughts, share knowledge, attend conferences, network .. think about WHERE to go deep (but dont go too deep on any topic yet) and start exploring

Wow them and Connect: Once you have a good portfolio online, start pinging important folks who may open new doors for you or challenge you pushing you into meaningful directions

What else is useful? Your thoughts? https://www.facebook.com/notes/ramesh-raskar/how-to-startfind-ideasget-guidanceconnect/10153816517413179?notif_t=like

rraskar11 karma

(a) You can look at some good tips on http://cameraculture.media.mit.edu/join-us/ about how to build your research and project portfolio. (b) We are always looking for multi-dimensional candidates for grad school as well as for faculty positions at MIT Media Lab. For grad school admissions, Grades are not the most important but if your grades are average, you can show amazing achievements in some other area. FYI, I was topper in HSC (12th grade) in Maharashtra but did not finish among the top at the end of four years of college. (c) There are many groups indeed. Maybe I will post later. (d) You can look up at posiitons on http://www.media.mit.edu for broad positions.

tempaccountin1 karma

Can you please also elaborate how important do you consider research papers to be while evaluating graduate applications? I have decent grades but no research papers.

rraskar3 karma

Greatly helps to show you have Publish->Demo->Deploy portfolio. But you can show research output in many ways: competitions, awards, blogs, opensource libraries etc.

brotherinarms19926 karma

Hi Dr Raskar,

Question 1: How has your research focus and approach changed over the years? i.e Young Individual Researcher Ramesh vs Professor Ramesh leading a lab.

Question 2: Tell us something you've never told anyone else before?

rraskar16 karma

Great question. I recently gave a talk at UIST in November about my own transition. Hopefully it will be online soon. http://uist.acm.org/uist2015/about

Transcript is here: https://docs.google.com/document/d/1fPqOknXvFNxDsc1LaTVl6W6L-lN_OwBnqAYqHUnGxcg/view

I feel that all of us go through four stages in life: LEARN, APPLY, IMPACT, ACTIVISM. First two are well understood. Impact: your idea is used by dozens, thousands or even millions of people. Constructive Activism: you inspire others via a message or a process.

But we ideally should go thorugh the four phases in that sequence. If you jump straight to impact or activism, your credibility is low and your efforts will have minimal effect. (I am ignoring distructive activism here .. it is of course easier to break or simply oppose things, but what matters when you are constructive)

mistermagicman3 karma

Hi Dr. Raskar,

In 2012, I had the incredible chance to be at the media lab for two weeks, and helped Christopher Barsi work on a UI for the femtosecond camera. I spoke with you a few times, but I believe you had a conference and you were pretty busy. Anyway, I had never really programmed, and I think I wrote about 2,000 lines of code just to make an array of buttons — I didn't know about looping yet! That experience inspired me though, and I'm now a junior majoring in Computer Science.

My question is — do you know if any of the code I wrote was ever used for anything? Thanks for the AMA!

rraskar2 karma

Glad to hear from you! Ping Chris.

360walkaway3 karma

So do you act like the guy in the Ajay Bhatt video a lot? Because I would if I had your credentials.

https://www.youtube.com/watch?v=YpiylFF6sV4

rraskar3 karma

Haha .. I have met Ajay a few times (they hired someone else to play his role though). https://en.wikipedia.org/wiki/Ajay_Bhatt

rikerw2 karma

Hi Dr Raskar,

I'm completely fascinated with some of your work and I found your TED talk to be exceptional. Do you have any summer internship positions for a third year undergraduate chemist, or know of any which are related to your work?

rraskar2 karma

samipjain2 karma

Hi Prof. Raskar,

What was the best experience you had from Kumbathon till now? I wish to participate some day.

rraskar3 karma

You can read about MIT Kumbhathon efforts at the page http://kumbha.og

Join the next Kumbhathon near Mumbai Jan 22-28th. Apply today!

rraskar2 karma

You can watch the videos of MIT involvement of Kumbhathon at https://www.youtube.com/user/cameraculturegroup/videos?sort=dd&view=0&flow=grid

samforthereddit2 karma

Hi Professor, want to know how you simultaneously manage top class research at your group along with many hackathons/workshops/huge projects like Kumbha Mela?

rraskar4 karma

Simple answer: via open collaboration. Our group at MIT are mainly scientists and majority of time is spent on research. We are just fortunate that we have been able to keep our platforms very open and invite thousands of like minded people to join http://lvpmitra.com (solutions for visually challenged), REDX in many parts of the world (http://redx.io), Kumbhathon (http://kumbha.org), Emerging World Special Interest Group (http://www.redx.io/emerging-wohrlds-1/#emerging-worlds). But you are welcome to take them forward in your own way.

physiologic2 karma

Hello, I'm a medical student who started a group at my school devoted to trying to address developing world health needs through technology. One of the difficult problems we have run into is that assessing the needs of the developing world requires extensive time spent there, and the information gained by such trips is not always well-shared between groups with these common goals. I was wondering if you know of any group repository or network by which we could connect a group, operating somewhat independently and geographically isolated, to other groups such as those operating in boston?

rraskar2 karma

Very good point. This is exactly why we started. Building network to identify problems can be cumbersome. And building sandboxes to deploy ideas can be even more challenging. So REDX chapters solve that problem. Our annual events are a great way to get started. http://redx.io

rraskar1 karma

From my UIST talk:

Question: You said that effective partnerships are critical. How do we form effective partnerships?

Ramesh: The partnerships are very time consuming, because by the time you build the chemistry, by the time you identify the right problems and solutions. And that's why we started the REDEx platform. Rethinking, engineering, design, and execution. Where this communities, emerging communities. And they're not necessarily in developing countries.

                                They could be in any emerging areas. Are already anxious to find these problems, to solve these problems. And so we have been very fortunate to have constant connections with this, and it takes a lot of time. It takes two, three, four years before you actually have a meaningful initiative that's up and running. But going back to your question, I think we need those avenues. Not just universities and labs, but labs all over the world, or the world itself as a lab, so we can go forward.

                                All the REDEx centers and partnerships we have are open to any of you. And that's the only way that it's going to go forward. So I agree with your partnership for challenging. But I think that's going to be the new model. Those are the new places for the organization and deployment.

More at https://docs.google.com/document/d/1fPqOknXvFNxDsc1LaTVl6W6L-lN_OwBnqAYqHUnGxcg/view

humanity_needs_fiber2 karma

Hello, Dr.Raskar. First I would like to sincerely extend my appreciation to you and your entire team. What you are doing hits home for me. Can you tell me if EyeNetra's current research goals includes any optical device that will help patients with Retinitis Pigmentosa and Glaucoma to see again? These are two eye diseases that currently do not have a cure, and can lead to permanent blindness. I have been diagnosed with both glaucoma and retinitis pigmentosa; blindness runs in our family. I am desperate to regain my eyesight and prevent any further degradation, therefore I am always very eager to hear about the latest in visual advancements. Thank you for all that you are doing. Losing your vision means losing your independence.

rraskar5 karma

Thanks for your important question. We have been obsessed with prediction of eye as well as other systemic conditions using eye selfie. And if collectively, we are all successful, we can indeed predict the onset of these conditions months of years in advance.

See our research at http://eyeselfie.org Plus my TEDMED talk at http://www.tedmed.com/speakers/show?id=102077&ref=talks How do we look at the future of health with both eyes

-notacanadian2 karma

If I wanted to make a coded aperture for a DSLR lens, what impact does focal length have on determining the best "code" for the aperture. Is there any commercial lenses available you would recommend using to start with?

rraskar3 karma

I recently gave a talk at ICCP 2015 and clarified that we should stop working on coded aperture for focus effects! (Thus negating my team's work in this area.). I also spoke about the lost decade of computational photography and how we have wasted too many years working on the wrong problems. I will try to post the transcript very soon. There were no slides.

rraskar5 karma

But here are other 'future of photography' articles I wrote or was interviewed.

http://www.popphoto.com/how-to/2009/08/what-photography-will-look-2060

http://www.technologyreview.com/article/413118/photo-future/

https://www.youtube.com/watch?v=1bGVKvXG4dU (This was an ambitious talk about future of Computational Photography in 2009)

-notacanadian1 karma

Thanks for the reply! I'm glad I didn't yet "dive in" and was just doing broad research. It's disappointing to hear that time was wasted, but I look forward to reviewing the transcript and where to look to the future.

Any tips on where I might instead focus my research on passive capture of both color/image quality and depth map in a single exposure? I've been focusing on capturing a static environment, with or without focus stacking to get both a depth map and the correctly exposed image as end products.

rraskar2 karma

Look at my co-author Amit Agrawal's page http://www.amitkagrawal.com/index.html

Our work shows tradeoff between photo quality and depth quality if you try to achieve both in a single photo.

So the simple answer is: 'Dont do it!', Use a stereo camera instead.

GeeveGeorge2 karma

Sir ,

1) What are you thoughts on implementing Machine Learning Algorithms to predict stampedes using real time cell tower data?

2) Google had recently open sourced "TensorFlow" their Artificial Intelligence Engine and it has really useful to me. Similarly , does some of the media lab projects get open sourced for the open community of developers to further work on?

rraskar3 karma

Geev very good to see you here. You are an inspiration to the next generation of innovators with your work. Yes using predictive analytics with cell phone tower data is a very powerful approach. We are working on this now at Nasik innovation center. Please apply at http://kumbha.org for our next meet Jan 22-28th.

JeetRaut1 karma

Hello Dr. Raskar, is it possible for for Non-resident Indians to apply for the REDX internships?

rraskar1 karma

yes. Submit online at redx.io

saikrishnagv1 karma

hello Prof.Raskar,

I attended your talk recently at IIIT-H. I just want to know how you are motivated to take up multiple projects at a time? How could you manage your time with too many projects at hand?

rraskar2 karma

Our broader projects worldwide are always with our collaborators and they are all amazing! (You do not hear about the not-so-amazing ones because those efforts slow down on their own :-). So the credit really goes to our partners, as they co-innovate with us and we learn a lot from those interactions, and apply it to our next platform.

At MIT we spend nearly all our time on research and innovation and solutions to deploy them. I define broad directions for research and share my juicy ideas we could work on, and guide research and often get into the trenches. But we have a flat hierarchy in the group and everyone has a lot of freedom to work on most pressing problems and very exciting algorithms/solutions within those broad areas. We also expect senior PhD students and every post-doc to spend 1/3rd of their time to work on projects that take us in a new direction.

nilaymokashi1 karma

Hello Prof. Raskar,

What plans do you have in expanding the Media Lab's initiative in India? Having been at Kumbhathon I feel it just takes a focused effort and dedicating time to think of solutions for day to day problems. Do you think starting starting a large number of such innovation groups in colleges will help?

I personally am very eager to start such a group/club in my college where a group meets everyday, thinks proactively about problems and has brain storming sessions. Can such a group get any kind of mentoring from Media Lab?

rraskar5 karma

Yes we realize that REDX chapters in many parts of the world can make a huge difference. Please type below your info and you can apply for the license to host REDX chapter at your organization. License is free but expects certain milestones. More details on REDX are at http://redx.io

Type your info as reply to this message and someone from our REDX team will get in touch with you.

rraskar5 karma

You can also Apply here

Atharva_Kimbahune1 karma

Hi Prof Raskar, First of all a big thank you for giving us an opportunity for interacting with you. I participated in ReDX at Hyderabad in 2015 and had an amazing first hand experience of collaborative learning. Wish to participate in more such events but since I am in engineering 1st year, have restrictions due to exams and submissions. Could we think of an distributed ReDX where is partcipants contribute from their places in their free time like on Sundays at a pre defined time, and host a web based collaborative session managed by mentors?

rraskar2 karma

Yes we are setting up REDX chapters in many parts of the world. Please submit info on http://redx.io and see one of the answers above. You can also Apply here

tempaccountin1 karma

Which groups in Media Lab deal with Virtual Reality? I am interested to apply to Media LAb but I'm not sure which groups should I apply to (I am open to other areas too but this has been a significant part of my research work till now).

rraskar2 karma

There are many groups combining digital and physical at MIT Media Lab. Look up following faculty and PI as a start: Pattie Maes, Hiroshi Ishii, Michael Bove, Joe Paradiso, Kevin Slavin. And there are many more.

nilaymokashi1 karma

Hello Prof. Raskar

  1. I would like to know how much does having a research idea strengthen one's application for graduate program at Media Lab's Camera Culture? (Given that the idea is relevant to the ongoing research in the group)

  2. Is there anything you would like to advice to stress on while writing statement of purpose for being a part of camera culture?

rraskar3 karma

Let me share my thoughts on the 'Future of Graduate Career' as I answered in my recent UIST keynote

Question: I was wondering if you have advice for graduate students as they make their way through this entrance career path. What would you tell them?

Ramesh: I think they're going through a very destructive transformation of a classic PhD. As you can see, to have a real world impact is not just about publishing or a demo. It's going to be more and more about deploy. To do that, we need to work in teams with rapid iterations. But you know a Master's or PhD has traditionally been a solo effort that's multi year.

And that is very very challenging. I think by the time you graduate, the stuff that you're thought is interesting to work on, maybe it has changed. Or maybe some parallel technologies have made that obsolete. The way we really need to work is through teams and very swift iterations. So the same way the value of publication is going down nowadays, because people like to watch videos. They like to watch even research videos.

I think we have this very disruptive change that's coming up in PhD careers as well. I would be very surprised if, a few years down the line, a PhD student was judged based on their thesis. I think they'll be judged more on whether they were able to demo and deploy. And if they work in teams, and not if you create one solution, but create a sequence of solutions that support your fundamental thesis. Your fundamental argument. So I think it's going to be very challenging going forward as a traditional PhD student, and so I know my advice has a sort of bias built into it and a lot of attribute. But I would say work in teams, be out there. Don't do here or there, do it together.

See more of the talk here: https://docs.google.com/document/d/1fPqOknXvFNxDsc1LaTVl6W6L-lN_OwBnqAYqHUnGxcg/view

Gulliver20112 karma

What's the purpose of a PhD then? Probably industry already does it well.

rraskar2 karma

PhD degree program as a learning experience is extremely valuable. It is learning about learning. But the notion that you should do a solo multi-year project is questionable.

mayurxy081 karma

Well I have more questions! If I have one idea which can really make a difference in world and don't know how to tackle it in some way. will MIT Media Lab be able to guide and help me to work on that idea ? What should be my approach in contacting MIT Media Lab and have suggestions on idea? Thank you

rraskar3 karma

See my Quora page https://www.quora.com/profile/Ramesh-Raskar for some answers.

ssarthak5981 karma

Hi Sir, I would like to know how was your life at school 20-25 years from now, and from there how did you reach MIT?

rraskar1 karma

Some details on my academic journey is in this story by Larry Hardesty

http://news.mit.edu/2011/profile-raskar-0929

ActsJuvenile1 karma

Dr Raskar, can you please ELI5 what are the main differences between computational photography vs simple stitching/stacking together of images? Why should I be excited about it?

rraskar2 karma

See my answer about 'Lost decade of Computational Photography' above. The 'burst mode photography' as useful as it is for overcoming camera limits, has misguided our field. See my talk about on real potential of computational photography.

hipsterdocmd1 karma

Do you have any plans for a non-mydriatic retinal camera -- or adapter for conventional iPhones or Android phones -- to complement the other EyeNetra mobile devices?

rraskar1 karma

See http://eyeselfie.org/ our latest work

phatskat1 karma

Where did the name Media Lab come from?

I worked on a project for Big Telecom Company and we originally dubbed it MediaLab - the lawyers said no, and now I can ask where you got the name from because we thought it was kinda neat.

rraskar1 karma

See my talk on What is 'Media' in MIT Media Lab http://www.slideshare.net/cameraculture/what-is-media-in-mit-media-lab-why-camera-culture But also curios to hear your thoughts

pessimish1 karma

You looking for interns in medical school over at EyeNetra? I'm interested in eye care innovation, but it's hard to get that kind of exposure in medical school.

rraskar1 karma

You can indeed write to vitor at eyenetra dot com or to EyeSelfie.org team at MIT