The HLF Portraits: Alexei Efros

Video thumbnail (Frame 0) Video thumbnail (Frame 1504) Video thumbnail (Frame 10053) Video thumbnail (Frame 21692) Video thumbnail (Frame 33331) Video thumbnail (Frame 35375) Video thumbnail (Frame 47409) Video thumbnail (Frame 59443) Video thumbnail (Frame 71477) Video thumbnail (Frame 83511)
Video in TIB AV-Portal: The HLF Portraits: Alexei Efros

Formal Metadata

The HLF Portraits: Alexei Efros
Title of Series
No Open Access License:
German copyright law applies. This film may be used for your own use but it may not be distributed via the internet or passed on to external parties.
Release Date

Content Metadata

Subject Area
The Heidelberg Laureate Forum Foundation presents the HLF Portraits: Alexei Efros; ACM Prize in Computing, 2016 Recipients of the ACM A.M. Turing Award, the ACM Prize in Computing, the Fields Medal and the Abel Prize in discussion with Marc Pachter, Director Emeritus National Portrait Gallery, Smithsonian Institute, about their lives, their research, their careers and the circumstances that led to the awards. Video interviews produced for the Heidelberg Laureate Forum Foundation by the Berlin photographer Peter Badge. The opinions expressed in this video do not necessarily reflect the views of the Heidelberg Laureate Forum Foundation or any other person or associated institution involved in the making and distribution of the video. Background: The Heidelberg Laureate Forum Foundation (HLFF) annually organizes the Heidelberg Laureate Forum (HLF), which is a networking event for mathematicians and computer scientists from all over the world. The HLFF was established and is funded by the German foundation the Klaus Tschira Stiftung (KTS), which promotes natural sciences, mathematics and computer science. The HLF is strongly supported by the award-granting institutions, the Association for Computing Machinery (ACM: ACM A.M. Turing Award, ACM Prize in Computing), the International Mathematical Union (IMU: Fields Medal, Nevanlinna Prize), and the Norwegian Academy of Science and Letters (DNVA: Abel Prize). The Scientific Partners of the HLFF are the Heidelberg Institute for Theoretical Studies (HITS) and Heidelberg University.

Related Material

Arm Multiplication sign Internet forum Normal (geometry) Musical ensemble Family
Point (geometry) Multiplication sign Hidden Markov model Device driver Student's t-test Mereology Machine vision Computer programming Formal language Wave packet Neuroinformatik Frequency Mathematics Different (Kate Ryan album) Bus (computing) Booting Boss Corporation Electric generator Inheritance (object-oriented programming) Gradient Content (media) Bit Connected space Data mining Personal computer Word Process (computing) Physicist Factory (trading post) Calculation Formal grammar Computer science Right angle Musical ensemble Whiteboard Quicksort Game theory Family
Point (geometry) Service (economics) Ferry Corsten Multiplication sign 1 (number) Letterpress printing Design by contract Computer programming Formal language Mathematics Goodness of fit Authorization Energy level Determinant Position operator Social class Standard deviation Cellular automaton Moment (mathematics) Bit Type theory Word Personal digital assistant Universe (mathematics) Video game Right angle Game theory Family Window
Point (geometry) Turing test Trail Group action Multiplication sign Direction (geometry) Control flow Student's t-test Mereology Computer programming Machine vision Field (computer science) Neuroinformatik Number 2 (number) Formal language Medical imaging Internetworking Term (mathematics) Robotics Computer hardware Circle Series (mathematics) Office suite Social class Area Boss Corporation Assembly language Artificial neural network Commutator Objekterkennung Physicalism Counting Library catalog Shareware Type theory Word Universe (mathematics) Computer science Video game Natural language Whiteboard Object (grammar) Quicksort Family
Pixel Group action Matter wave Multiplication sign Decision theory Direction (geometry) Mereology Neuroinformatik Medical imaging Coefficient of determination Semiconductor memory Analogy Videoconferencing Office suite Series (mathematics) Position operator Thumbnail Algorithm Email Texture mapping Gradient Demoscene Connected space Type theory Tower Textursynthese Computer science Right angle Resultant Thomas Bayes Row (database) Geometry Point (geometry) Color management Student's t-test Machine vision Hypothesis 2 (number) Wave packet Frequency Term (mathematics) Nichtlineares Gleichungssystem Analytic continuation Planning Cartesian coordinate system Computer graphics (computer science) Visualization (computer graphics) Personal digital assistant Universe (mathematics) Speech synthesis Video game Table (information) Family Window
Internet forum Bit rate
professor let's go back to way before you were professor Oh guys go back to your childhood in Russia and I'd like to know where you are as a child and what your family is like well I was I was born and raised in st. Petersburg which was then called Leningrad in Russia which was then called the USS arms a lot of things have changed and I was I was raised in a you know absolutely normal family normal for you know Soviet times
in an apartment in in but not normal in the sense of it being filled with science as a culture that's true so my my my father is a physicist my grandma and grandpa were chemists and and so indeed this is a disposition of the family yes yes so my uncle was a physicist as well so there's definitely was a lot of scientists around me when I was growing up also I would guess the quality of schools in st. Petersburg Leningrad was quite hot well yes and no so I think I think in general you know education and technical subjects at least was much higher in in the Soviet Union than that in the West we just think that they drilled much more math and to us and these these were definitely things that that even in they're not very good school you just you you had to learn much more but I actually went to it it's kind of weird so I I went to a special boarding school for low vision children because I was born with with a visual impairment reasonably severe and and and so and also at that point there were no real visual aids so by the time I got to America it turns out that there are special glasses that you can wear that you can read things and there is a special little molecules that you can see the board but back back when I was four you know in a kid in Russia there was no such things there and so I think my parents were worried that I might not do well in a normal classroom yes and probably they were also worried that I would just get beaten up yeah by by this you know the poet yeah because you know it's it's it's it's a tough environment and if you don't see well then it's it makes it even tougher and so and so they placed me in this special special school for for low vision children where you know you'd still get bullet but you get bullied by also half blind students so it wasn't so bad they would miss half the time yes yes so because st. Petersburg is a big city you know five six million people and the the school was on the other age of town and so it was an hour getting there and so that's why it was a boarding school most students would go there and and stay the whole week or maybe half the week and but I didn't I didn't really like that this whole institution yeah the institution for leaving you the family of the week or all of the above everything yeah and just but in general just I mean if you can imagine you know this institutional you know Soviet you know there is a lot of a lot of bad food and a lot of communism all the time and strict discipline and so I I didn't like it much at all and so actually I think I started commuting home starting it like h97 seven so starts at seven yeah so I think after about two or three years I started actually commuting you know fifteen minute walk fifteen a minute bus 15-minute metro or than fifteen minutes walk again but in the end I actually enjoyed that and especially like the metro you're not going going into there on the train I do Mike about a little baby brother okay so yeah I'm the oldest is a big question you can answer it anyway what are your parents ambitions for you job when do you begin to feel that I I think my parents didn't push me that much I think partly who they were worried that because I don't see well you know I should be I should be given some broader broader broader leash or longer leash and do things my own way also I think the generation of my parents especially in the sciences they were I think now it's called the the the tiger parents you know there was definitely something like that and when I was growing up in the Soviet Union where the parents would really push their kids and make them you know learn languages and learn music and learn you know this and that and the other and and I think my dad was actually kind of was
against that he felt that that kids should just be kids and not be pushed around he was in early America yes I think so he knew it I think no I think he was definitely he was there was definitely you know he read a lot of healing he knew English as he read a lot of kind of American American literature and he liked it and he he felt like yeah you know just let that the kids be kids something will work out and so I wasn't really pushed around I wasn't like in fact I you know I didn't know what learn how to read until I went to school which is completely insane you know you don't learn until you know you're seven yeah yeah but he felt that you know it just it's okay so yeah no I I I think there was definite I wasn't a good student I was I wasn't the bad student but you know yeah so getting like Beason in math for example my my felt that my father was she wasn't particularly upset here but he was like you know just quizzical like hmm hmm you know right and you know and you know I I I did I wasn't dad I was a bit lazy so so you're not the first among these interviews scientists right that about the childhood but you know in a way it suggests an ability to think in a free-roaming way rather than just directed by maybe that will help you one day is a research scientist but yeah anyway as a child you're not being pushed I wasn't special in any way at all maybe if I had some sort of a kind of thing that I did well was I wrote good essays I would always get very good marks on my essays for literary content they were all very literary and I but I would also always get the worst marks for the further grammar so I would never get a very good grade because you know the content was great but the grammar I would make you know three mistakes in every septum everywhere in every word so you're still there Russia I know you you'll leave at 14 and we will talk about that but er until 14 is there a point you take fire intellectually or you begin to read a certain kind of book or you have a teacher that inspires you is there anything in this period I think frankly I think I actually you know I I formed this individual by the time I left Russia at 14 I think okay so I did after that it was all kind of going downhill from there really no I think though yeah I think there was a couple of things the one is that I I the one thing that that my parents did tried to make me do was to they you know made me take the adolescence okay and and the thing there was that I just really liked it and I was lazier practicing and I was never any good but I just really liked it and I um I you know I still play the piano not very well but I enjoy it and you know I would I started to compose and and just really started to appreciate and enjoy music and that was a there was something that you know made me happy and it was something that I was you know kind of discovered of my own might know nobody in my family was musical at all hmm I was told that my grand uncle's were for a very musical but I never household not really no yeah so it was like one of those things that was mine that I enjoyed and the other big thing that happened was that my father got interested in in computer science as a physicist it was something that he was very curious because it could solve some of his problem so he was he read about it of course the nobody fighting computer but but you know he had a programming calculator and he was so excited about it that he would like show this to me and and he would teach me he taught me how to program the calculator and and I really liked it and and then and then and then he went - he was pretty straight that happen and he was allowed to go abroad for for for conferences and I think from from Germany he brought in one of those in HP programming calculators it was magical and there was all this there was this manual holy German none of us knew German but like I figured out what how the program is in German without knowing any German and and this would have been one of the few in Russia yeah there wasn't that many yeah that's right that's right that's right and and and and and I remember and then I really got into this whole come yeah I got into this oh and then and then and then my father met I think it was like he he met in the subway his high school friend or the the guy he went to high school at after 30 years and I said oh yes oh good to run into you I you know my driver was sick so I didn't take the car this this time but I'm taking the subway so good to run into you turns out that he is a big shot you know party boss and the part and he was it was a he was in charge of making you know rubber boots he was a chairman of the rubber boot factory and then the party sent him to do a different factory and that Factory was baking I'm so v2 union's first personal computers you know he didn't know anything about computers but there he was in charge of a factory and in you know in the Soviet Union if if you if you're in charge of something you know you you share with your buddies hey you know I'm I could have offered you rubber boots last year but now I I don't have access to that but I have this personal computers you want one and my dad yeah yeah let's do it and and so I got one of the first Soviet made personal computers in my home through this weird connection I think maybe twelve I'd say I think yeah I think it well and and it it was a revelation it was you know I didn't have any he didn't have any programs they didn't have any games it had nothing and and it it was also you would it would overheat so after an hour he would crash so not only did I had to learn how to program it do it to make it do anything interesting but I also had to do it really really fast because I had to code the the program also you couldn't save
anything to do there was no there was no way to save anything so you're talking about the dark ages yes so so you turn it on you type something in then you know whatever you broke your program at you and you play with the program maybe you'd you know program a game you play with the game but then after an hour or so it would freeze up and you know you have to start over again okay time to bring you to America uh-huh what brings your family to America um so my my father was always a little bit of a dissident so he was you know he was he would go to read some you know forbidden books go to some forbidden art exhibitions and in general just would not cooperate with with that with the party authorities and so he was not really given he was not really given much both gives his progress and in in in it was was limited he was not here I remember he would get these invitations every year who get in really you know you know I would love to come you know invite you to Canberra to Australia for a semester as a visiting you know professor all-expenses-paid and of course he couldn't go and it wouldn't let it say you know they would not let him go because you know you had to have an exit visa you had to have an how you will should be allowed to to get out of the contract so he was a pretty you know famous scientist at that point but listen Sam he was definitely not sound and so I think they were thinking of leaving earlier you know when I was really young and then but that didn't quite work out and then I think that there was kind of this idea that if there was an opportunity that that might be something that that would be that that might be a I think I think my father was in there in the back of his mind there was always the Germany in the late thirties there was a lot of good scientists who you know worth realizing things are not good and what they were thinking that maybe they should leave Germany you know especially especially the Jewish ones but then they were kind of not quite sure then by the time they realized okay this is time it was too lately yes and things were happening that were very changes were very fast and some were good but some were very scary the level of anti-semitism for example in the late 80s has really low rise and risen up it could have gone different ways I think in the end nothing really terrible happened in in the Soviet Union of broke up in a reasonably reasonably controlled way but it could have been much worse they could have been some this was that early late eighties yes when he was coming I think that's right that's right and so and so he felt that you know and and the window of opportunity might have also closed so nobody really knew like nobody expected this chance that suddenly you could leave right and it wasn't clear how long that window would stay open and my father felt that he will not make the mistake that those Germans did in the thirties who did not get out in time so he know so he what happened was that he had a good very good colleague and friend that UC Riverside who it was still not very easy to leave so he his he was this friend of his arranged for him a three-year visiting professorship position because if only if you leave for three years were you able to take your family you still had to get a approval from the you know the the president of the Soviet Academy of Sciences but you if you anything less than three years you leave your family behind but if you could get something for three years they would they might allow your family to go with you and so he got this very extraordinary three year visiting position and so we went to do it to California with that former Soviet resin right right now in America in California yeah he's put in school yeah how's his English I knew I knew one word and then it turns out that it was actually two words ice cream so I I doubled my Cavallari you know in the first day of class that was pretty good it it was it was stuff it you know 14 year old I just I just started making friends in Russia you know it was kind of a lonely geeky kid growing up so it was I was finally happy to have made friends and get my no no my cell my way around and travel around by myself and then suddenly I get placed in this new weird world with a language I don't understand also it was a big change between a hanger six million City you know quite for cosmopolitan with a lot of you know subways and whatever to a tiny little Southern California town with no sidewalks and so it was it was a tough adjustment I we're always looking work you know where where is the city where we're going to the city and this is it this was the city just this this you know Rose you know just one one story houses as far as the eye could see how are we gonna get you as a fourteen-year-old knowing two words in English to even qualify to go to university in the nineties four years is there some magical moment is it a new determination to learn no I just I I think this is just your standard I mean it's a standard immigrant thing you know that you just have to figure this out it's not I know and my family also the same thing or my father in you some English but the rest of my family we didn't know anything and you know we needed to somehow figure this out and and and you know it it didn't seem like a big deal because you know it was reality it's reality I'd like it in the
Soviet Union right life is tough so this was just continuation of that you know you just have to do it and some things were were really quite amazing i-i-i was not happy with america i was not happy with the fact that you know that that there was there was you know everybody was driving and I couldn't drive and so but there was some things that really were absolutely magical they my my dad mentioned in the school that you know I don't see well and suddenly all of this you know assistive services people is oh oh okay we will assign this this this this person to you sorry I I guess I still remember mrs. Simpson mrs. Simpson from from the you know from what is it handicap services or something big woman with a huge smile you know in Russia people don't smile right so there was like I I it was a big thing for me just this huge smile and she's like okay we will figure this out don't worry and so the first thing she did was she got me glasses where I could actually read normal print for the first time in my life so that was quite something then she got me a spyglass so I could
see the board also that's something I never managed to do to see before and then she would get you know she she took my little Russian English dictionary and she blew it up so it was like this big so I had to carry it within a backpack but you know the letters were this big but I could actually see it and and she taught me like you know how to cross the street and were Dukan of how to catch bosses and how to find you know house numbers and she was just so optimistic and so positive and it just it was so good and it was it was really something I fell you know this this country you know they usually less forward it was I think it was I think it was yeah I have get you to the point where I will begin to believe they're going to be a scientist what does it take to get you there I I don't think they're I think it's all just a series of accidents I don't think there is any it could be accidents yeah but there still has to be a time when first of all you have to decide in the university you will write learned at some point you have to pick a major yeah you have to begin to shape your future how does that happen I so I think so I was you know since the time I got this computer back and I was I was I was hooked I would just I would do I would do I remember a back even back in back in Russia yeah you know you go I got sick a lot so you know you get sick it's great because you don't go to school which I hated but then for two weeks you don't have to do P you don't have to do physical education so I remember I was sitting there everybody else was running like crazy around in circles and I was sitting there relaxing and reading this book about the pdp-11 assembly language which is the most boring weird strange thing I don't remember why I was reading it big I you know I didn't have a PDP a computer so it was completely you know academic I just loved it I'd love to ideal it and so um and and yeah even in Russia I you know I I would i coded up some crazy thing and they send it out and I you know do some contests and I got I think second prize for Lincoln this whole Soviet Union you know programming contest yeah yeah so so there was you know I was doing that and then going to America again I you know didn't have the English didn't have many friends but I had this amazing computer well so the family obviously got the latest computer once you arrived no not really I think I was like I would actually go to my dad's office and I would just play around with his computer okay and yeah I would totally one of was his computer he couldn't use it at all because I was on it but I and I will I I was writing I was coding things I actually even wrote like a program that I would give out a shareware that gave it out to put it on online and then yeah something like that yeah yeah and then I would say okay you know if you liked it please send me 20 or send me 10 and a postcard from me where you live so I got a lot of postcards and I get you know I got somebody to give people like five 600 noticing this ability or direction yeah I'm just wishing I was just having fun and I think I think if I had if I stayed in Russia and I have had my friends and other things I wouldn't have been so focused on on computers it was really just that I was all by myself and I didn't have anyone to talk to it so I think it might be something like that like EXO dorm unlike I don't know who knows maybe it would be better if I had spend my my teen years hanging out with friends so I'm gonna put you in University you talk probably because your father moved there yeah so exactly so my father moved to Utah and it was you know you know I I applied I applied to various universities but you know it did it was it was kind of yeah so I applied to for example in you know Carnegie Mellon ears too which is famous for computer science but then we looked at how much it costs it's like one of those like is that is that a phone number is that the price you know and then you know university of utah was a great deal it was something like you know five hundred dollars per semester or something so economics that you did modely and it was it was a fine it was a fine University and and also I while in high school I would they had a very nice summer program which was which was wonderful that I took and there were they also held programming contests and also in high school here in new talk show that I had a very nice computer science teacher so in in this high school there was actually a computer science oh.he in the high school and so and the computer and the teacher realized that I already knew everything and so she just let me just play around with all their equipment and so I I you know I put the school on the internet and I was like assisted ministering their there their computers and and and she just let me do whatever I wanted mrs. green she was great story of your life people are letting you do pretty much yeah what you want yeah it was wonderful I just they just just you know whatever do whatever yeah chant University which is a economic market is where you go here in Utah but when do you find this special well you have to choose a major it's um right no no I always knew I wanted to come at that point it was clear I want to do computer science and and and yeah it was up and you know and the first day of classes I remember distinctly I had you know two computer science classes the first day of University and I just after the in the end of the first day I'm like I can't believe this is so amazing this is exactly what I I learned so much in just one day and it's just exactly what I wanted to do is yeah yeah yeah so I've finished finished high school in 93 I think and so yes mid-90s so where are we in commute computer knowledge at this point it's a big question but I mean in general now that you're entering a more direct in my professional path what is known and what is making you curious well so I actually another thing what that happened is that because of there was nothing else to learning in high school even when I was actually finished happen high school beer or two I decided I'm just gonna go to the local university and see maybe they'll give me something to do and it was it was quite something I was I was quite a weird but kind of brazen kid at that point I guess because I just you know I went to the computer science department I went to the main office and I come in and I say take me to your chairman you know this is Russian accent
and you know they they take me to the chair of the department I say I want to do research and bless his heart Tom Henderson who was the chair at the time he's like okay weird Russian kid sure you know I don't care you're still in high school yeah why I you know you can do this this or yeah or I have this I got this old robot sitting around wide you know you want to play with that you can play with that right and and yeah and again it he just he let me do stuff he would meet with me every week it was absolutely absolutely fantastic so I'm gonna call him your first mentor I guess so I guess well I guess my father was really nice Center on explaining to me how these things are and then and then yeah and then and then and then tom was ya do ya let again let me let me play around with things there was a robot the robot would go around and you know I'd make it pick up trash we even you know send it to some competition at Seattle and you know we did reasonably well there so yes sir by that time I was already interested in in this in this direction and the transition to university is virtually done I mean you're yeah you're already technically it was yeah I didn't know anything but I was already I already knew people so that was kind of it must make sense to just just hang around and like all the graduate students II knew me already and I was I wasn't know in the count on campus they all the time would eat all the free food they would is the is the term artificial intelligence beginning to circulate around you at this point or not an important concept yeah um for you are available delicious was definitely something that I was very interested in from the very beginning at that point in time midnight is it was a it was a bad word artificial intelligence was something that people tried and failed I was still very much interested in the idea and that's kind of what got me to computer vision because I thought you know why was artificial intelligence not not so much was kind of you know it was a a I've winter at that time and one of the reasons that I could see was that there was no real way to figure out if it was doing anything and so I thought okay maybe we need to have some sort of M type of AI where you could really see if it's doing something and so and so you know natural language seems like one of those things you know the Turing test so you know you if if a computer can answer questions for you or it can converse with you then you know it's actually really working but then I thought okay no this is this is too complicated this is too hard then robotics seems it like another thing that okay if you can get drawn but to do something for you then you know it's actually working but again robotics you know or you have to deal with Hardware it seems also Hardware breaks all the time and so then I could have realized that you know vision it's it's something that it's it's hard to fool you know you you know give computer in image and in computer needs to be reason about the image and and find objects for example it it's if it's actually can do that that you know we know we it's actually doing something hard to fool it's not as hard as language we know that you know most animals have vision only us have language and also I felt like being poor sided myself I have an insight because given how bad my vision is I shouldn't I shouldn't be able to you know deal with the world as well as I do and the fact that I can actually still you know navigate and get get plump little from place to place recognize people suggests that there's something interesting going on and that because kind of in my brain that interesting is kind of slow enough that I can almost sense what it's doing I said how maybe I have a kind of an inside track there and that's why you know early on I started getting interested in computer vision even yeah yeah beginning of undergrad yes yes yes that's right and and you know yeah we'll get your graduates before resume it still does an undergraduate it's probably a word you'll you'll resist but you see you're precocious possibly in setting a field already as an undergraduate and in fact it's where you will do your work yep yeah that's kind of unusual angry angry are you getting the both the technology to work with that is helping you advance these ideas are you getting teachers that are helping you in computer vision questions or you again pretty much doing it by yourself oh no I think I mean first I'm not really doing much myself I'm still just learning right okay I mean I'm coding some silly things but it's I'm not advancing science in any way no I there's definitely been even actually thinking back you know my father back when I was nine he was telling you oh you know there's this new interesting area rather than Indian Scientific American called you know object recognition and that it's it's very neat because you know you get the computer to recognize objects for you and you know and you for some reason he thought it was connected to physics so he thought it was part of physics but he was telling it to me and so I you know I got thinking about this early on but then Tom Henderson this theater he was a he was a he was computer vision a researcher and Bill Thompson and other computer vision research earth basically they let me play around when in their lab and and and they you know they taught me a lot of things and their graduate students too and so I definitely I got immersed in that culture and I you know I I was aware of you know what things were happening and aware that you know basically not much was happening really it was it was you know it was this feel that was still kind of waiting for its heyday but but I was definitely by the end of my undergraduate it was clear that computer vision was was something that it was interested in okay now you have to decide and graduate school yeah where are you gonna go um so I wasn't I definitely wanted to go to graduate school because I felt that though I basically took all the classes I could from University Weaver including graduates classes so basically I think I basically did the whole catalog which is what you did in high school on the way to college kind of yeah and still I thought that
would be more things that I wanted to her so I wanted to to to to to go to graduate school but actually I didn't want to get a PhD I felt that I wasn't good enough to be a scientist but I just wanted to learn more stuff and then I could go and and you know apply it somewhere but of if you if you do uh if you do masters you have to pay yourself or you know for it yourself but if you get it if you apply for a PhD you can get a fellowship economics again exactly so I trickily applied for a PhD but because I was sure that you know I wouldn't I wasn't good enough I applied to 20 places 20 yeah send out 20 applications strangely got to pretty much all of them very well I was very surprised by that and and then yeah and then and then for you know various reasons ended up in in Berkeley would you have said at that point of course Berkeley is very well positioned now in the developing computer yes yes but is it particularly well positioned in computer vision issues um now it is but back back in the day you know my my who you know the person who became my advisor Jitender Malek me he was he was actually it was a young kid by then you know I when I asked my undergraduate advisors it you know you can go to Berlin and play around with this young kid Jitendra and I realize now yeah he is he was younger than I am now what-what I joined his lab so he was a young kid now he's like the the most important person in our field but back then he was just starting really I don't know I just it wasn't it was again kind of just a very natural decision not not very scientific I I didn't I didn't feel like MIT was I didn't feel like the atmosphere of MIT or Stanford it was was was that good and you know and CMU was a great place but but I wasn't that excited about it the tower so you'll get the revenge yeah exactly what for good exactly dirty you know the bay the the heels it was nice yeah ya know because I've read about you I know that the next question for you should not be what was your thesis because you start doing work that gets noticed and advances the peel while the graduates right can you tell me about a couple of those insights and the papers that you were beginning to present well again I I I really think that my thesis was basically a failure in that I I got I got a I got very lucky in the very beginning and I came up with a very very simple just trivially simple algorithm to synthesize visual texture so you have a piece of visual texture and you want to create more of it and it turned out that something that I read back when I was a kid in Russia about synthesizing natural sounding text turned out that using that as an analogy to pixels I could just have a very simple way of synthesizing good looking textures and so and that I did my second year in grad school and that year is what 98 98 99 years and and and that that became very famous and and because the results were just so much better even though that this algorithm was just absolutely trivial it's the term this is a well this is basically this that this idea of you could say that it's basically the idea of using data-driven methods so instead of instead of some of how the stealing information from from from the data and that kind of synthesizing from it you basically just copy pieces from the data it's kind of I I call it you know post-modernism in computer science writer you can just copy what people have or what what other in previous images have had have had and put put it in a different way and that there you go it's a very very simple idea it's just that nobody has tried it and and the results were surprisingly good and so that people got really famous and and then after that I basically didn't do anything for three years because you know how do you top that you know and so I was actually very stressed out because I was trying to do something really cool and nothing was working so actually most of my PhD I was very stressed out because like I felt like I needed to talk my first paper and it was it was very hard what was the HD on however disappointing it may be to you well so it was basically you know that first paper is plus kind of some continuation that for for for Vinick applying similar approaches to human emotion because you know my adviser told me you know by it that by the you know by the end of my fifth year my advisor said okay you know it maybe it's time to graduate don't worry don't try to you know do something really great you know remember PhD is really there you know it's the beginning without the ass yeah yeah exactly exactly by that time of course I also kind of realized okay maybe I don't want to leave with a masters maybe I want to stay around you know I had very nice office mates the adviser was very nice you know I had a seat next to the window so again I was reason you know I was thinking about going off to to industry at that point and then I laziness kind of kicked in and I just said I'll just I'll just try to finish the speech now let's pause on this question of the decision of a career in industry or academia how do you make that decision again you stumble into it yeah and said that you decide I want to stay I might as well do a PhD yeah I will did you think you might then go to industry or were you now thinking okay and that could Evert career is no I I was not sure I really didn't know like yeah I once I decided that I might as well stay around for a paged yeah I will still not had no no real plans for the future and I know a lot of people including some of my students they really have like a long-term 20-year plan and I might time I was definitely not that good i really had no plan at all and things just kind of happen you know i after after this paper that i presented on texture synthesis you know one you know famous professor came up to me and say said oh i really like your work would you like to do a postdoc in my lab in oxford and I'm like well you know this is only my second year so I guess you know I you know not now but maybe eventually but then I remembered
this and then by the time I was finishing I emailed this professor under the sermon I said remember three years ago you offered me opposed the position in your lab well how about now and to his credit Andrew I would adapt to it even though you know it was much later than he expected and and he said yeah sure come and for me it was just you know going to England Oxford beer it sounded really fun so intellectually not to interrupt the Beeman but intellectually it was actually also a great place to be at that lab at that time because at that point they were just starting to con transition from the traditional things that they were doing in geometry to object ignition and also doing a lot of this kind of data-driven things that I was also very interested in and so it was actually in retrospect just a perfect place to go to kind of soak up that atmosphere and also a lot of very good people were in under the sermons lab at that time and I got to meet them and and and it was it was I mean it was really even if at Berkeley it was a golden time to be at Berkeley so again I was super likely to be at Berkeley when I was because a lot of my PhD brothers older PC brothers worse were doing some really fantastic things and learning from them has been just super super useful for me and it was a gamut of this this golden era and then going to Oxford when people like Joseph Civic was there and Marc Evan Hammond and and Fitzgibbon all of them would just happen to be in that lab in Oxford and and again I was just very lucky to do so cop that atmosphere but you don't stay up for a year yes and then you go to earth well yeah so then while all of that was happening again I was you know I wasn't sure what I was going to do after the pause dog because okay I go to England that was my you know one year plan go to England and try the beer and hang out with the with the folks there and then I get another email that says oh by the way you know you I hear you might be graduating would you like to apply to Carnegie Mellon to for a professorship and this was again completely eye-opening for me because I never thought of myself as a professor at all but yeah if somebody is asking me why not and so again I you know I applied for for for a tenure-track assistant professorship at various universities and and and and yeah and got got got an offer from will you be there nine years nine years is this the period when just the phrase big data is beginning to circulate yes something like that so when I was I think my PhD was maybe one of the first that used dated driven methods which was kind of precursor to big data big data that kind of happened maybe five eight years later but yes something like that well your record of you melon that's right I tracked the Train I've interviewed other people in the series who who thought of big data in terms of text Mike Butler thinking in terms of image exactly exactly yes and so that that we were yeah we were pushing for images to be to be the really the big data because okay text this text but text you know you take all of the QPD you can you can basically store it on a thumb drive right with with with images or and videos you know it's that's really the big data and and at that point in computer vision and computer graphics people were still not very open to that direction and so again it was very lucky that I got to meet some a few people that were kind of on my wavelength one you know some people in Oxford like Joseph Civic and also I proved through for them I met another wonderful researcher Antonio Alba from MIT and the kind of the few of us were like yeah you know we need we need to get the data in the data into this because there isn't time to develop this fully would you mind setting the table in terms of where we are and you're a major participant in getting us here where we are in terms of what we know about big data and the visual and what the applications are starting to be just a kind of so I think I think the big story in computer vision you know for the first 30 years was really the idea that we needed to come up with algorithms that would go from the pixels to understanding and a lot of effort and a lot of you know sweat and toil has been spent on that story and something that a few one of us were were starting to understand in in you know early 2000s was that the data was a very important crucial part of the equation that it was maybe even more important than the algorithms the fact that when you're looking at something when you're looking at a scene it's not just the pixels of that scene that you're seeing it's not just this particular scene it's also that you're making connections to all the other scenes of that type that you have seen in your life so your your previous experience your memory the data that you have been exposed to you know for whatever twenty thirty years of your life all comes into play and that it might be because of this that that computers are so so find this problem so hard because the computers come to this as in with tabula rasa you know they don't know anything they haven't lived around they haven't seen all the data so for them this is just completely new and so the few of us were saying you know if we can just give the computer as much visual experience as we give to human kids maybe the problem will become much simpler and and indeed it turned out that this this was correct that that a lot of problems become much much simpler if if you use a lot of data and also algorithms turn out to not need to be that complicated anymore because the data can have takes over and the data tells you the right answer