Ari Shapiro

shapiroari (at) yahoo (dot) com
 
What I do right now...

Currently working as a Research Scientist at the USC Institute for Creative Technologies and lead the Character Animation and Simulation group.
2007-2010 Graphics Scientist at Rhythm & Hues Studios.
2006-2007 R & D at Industrial Light & Magic and LucasArts.

Ph.D. Computer Science, UCLA 2007
M.S. Computer Science, UCLA 2002

Caught up in the dot-com madness 1993 - 2001
Animation Software
SmartBody is a character animation system that gives an interactive character an extensive set of cpabilities and behaviors, such as: locomotion, steering, object manipulation, speech synthesis, emotional expression, gesturing, physical simulation, gazing among others.

The DANCE software is used for physics-based animation research, including dynamic simulation of rigid bodies, motion capture and dynamic control.


Research
You can access my Google Scholar page here.
You can also browse some of my older PhD
research in motion capture, physical simulation, animation and artificial intelligence.

Papers
 E. Miguel, A. Feng, A. Shapiro, Towards Cloth-Manipulating Characters, The 27th Conference on Computer Animation and Social Agents, Houston, TX, May 26-28
 
 A. Shapiro, A. W. Feng, R. Wang, G. Medioni, H. Li, M. Bolas, E. Suma, Rapid Avatar Capture Using Commodity Sensors, The 27th Conference on Computer Animation and Social Agents, Houston, TX, May 26-28 (video)
 
 A. W. Feng, Y. Huang, Y. Xu, A. Shapiro, Fast, Automatic Character Animation Pipelines, Journal of Visualisation & Computer Animation (paper preprint, bibtex)
 
 Y. Xu, A. W. Feng, S. Marsella, A. Shapiro, A Practical And Configurable Lip Sync Method for Games, ACM SIGGRAPH Conference on Motion in Games, Dublin, Ireland, November 2013 (paper, video, bibtex)
 
 S. Marsella, A. Shapiro, A. W. Feng, Y. Xu, M. Lhommet, S. Scherer,Towards Higher Quality Character Performance in Previz, Digital Production Symposium, Anaheim, CA July 2013 (paper, bibtex)
 
 S. Marsella, Y. Xu, A. W. Feng, M. Lhommet, S. Scherer, A. Shapiro, Virtual Character Performance from Speech, Symposium on Computer Animation, Anaheim, CA July 2013 (paper, video, bibtex)
 
 A. Shapiro, A. W. Feng, The Case for Physics Visualization in an Animator's Toolset, 8th International Conference on Computer Graphics Theory and Applications, Barcelona, Spain, February, 2013 (pdf, bibtex)
 
 A. Feng, Y. Huang, Y. Xu, A. Shapiro, Automating the Transfer of a Generic Set of Behaviors Onto a Virtual Character, The Fifth international conference on Motion in Games, Rennes, France, November, 2012 (pdf, video, bibtex) Best Paper award!
 
 A. Feng, Y. Huang, M. Kallmann, A. Shapiro, An Analysis of Motion Blending Techniques, The Fifth international conference on Motion in Games, Rennes, France, November, 2012 (pdf, video1, video2, bibtex)
 
 A. Feng, Y. Xu, A. Shapiro, An Example-Based Motion Synthesis Technique for Locomotion and Object Manipulation, Symposium of Interactive 3D Graphics and Games, Costa Mesa, CA, March 2012 (pdf, video, bibtex)
 
 A. Shapiro, Building a Character Animation System, Invited Talk, Motion in Games, 2011 (pdf, bibtex)
 
 Welbergen van, H. and Xu, Yuyu and Thiebaux, M. and Feng, WW and Fu, J. and Reidsma, D. and Shapiro, A., Demonstrating and Testing the BML Compliance of BML Realizers, IVA 2011, (pdf, bibtex)
 
 A. Shapiro, S.H. Lee, Practical Character Physics For Animators, IEEE Computer Graphics and Applications, July/August 2011 (pdf, video, bibtex)
 
 B. Allen, D. Chu, A. Shapiro, P. Faloutsos, On Beat! Timing and Tension for Dynamic Characters, ACM SIGGRAPH/Eurographics Symposium on Computer Animation (SCA), ACM Press, August, 2007 (pdf, video, bibtex).
 
 A. Shapiro, D. Chu, B. Allen, P. Faloutsos, The Dynamic Controller Toolkit, The 2nd Annual ACM SIGGRAPH Sandbox Symposium on Videogames, San Diego, CA, August, 2007 (pdf, videos), (bibtex)
 
 A. Shapiro, M. Kallmann, P. Faloutsos, Interactive Motion Correction and Object Manipulation, Symposium on Interactive 3D Graphics and Games, Seattle, Washington, April, 2007 (pdf, videos), (bibtex)
 
 A. Shapiro, Y. Cao, P. Faloutsos, Style Components, Graphics Interface 2006, Quebec City, Quebec, Canada, June, 2006.
(pdf, videos), (bibtex)
 
 A. Shapiro, P. Faloutsos, V. Ng-Thow-Hing, Dynamic Animation and Control Environment, Graphics Interface 2005, p. 61-70, Victoria, British Columbia, Canada, May, 2005.
(pdf, software), (bibtex)
 
  A. Shapiro, F. Pighin, P. Faloutsos, Hybrid Control For Interactive Character Animation, The Eleventh Pacific Conference on Computer Graphics and Applications, p. 455-460, Canmore, Alberta, Canada, October, 2003.
(pdf, videos) (bibtex)
 
  A. Shapiro, G. Fuchs, R. Levinson, Learning a Game Strategy Using Pattern-Weights and Self-Play, Third International Conference on Computers and Games, p. 42-60, Edmonton, Canada, July, 2002.
(pdf, Springer-Verlag, from LNCS) (bibtex)


Refereed Conference Presentations
 
  A. Shapiro, S.H. Lee, Practical Character Physics For Animators, SIGGRAPH 2009 Talk, New Orleans, LA, August 2009
(pdf)
 
  J. Bayever, J. Gordon, G. McMillan, Y. Lakhani, J. Mancewicz, A. Shapiro, Making Statues Move, SIGGRAPH 2008 Talk, Los Angeles, CA, August 2008

(pdf)
  A. Shapiro, P. Faloutsos, Interactive and Reactive Control, SIGGRAPH 2005 Sketches, Los Angeles, CA, August 2005
(pdf, demo)
 
  A. Shapiro, Y. Cao, P. Faloutsos, Interactive Motion Decomposition, SIGGRAPH 2004 Sketches, Los Angeles, CA, August 2004
(pdf)
 
  A. Shapiro, P. Faloutsos, Complex Character Animation that Combines Kinematic and Dynamic Control, SIGGRAPH 2003 Sketches & Applications, San Diego, CA, July 2003.
(pdf)


Refereed Posters & Demos
  A. Shapiro, A. Feng, R. Wang, G. Medioni, E. Suma, Automatic Acquisition and Animation of Virtual Avatars, IEEE VR 2014, Minnesota, March 2014
 
  A. Leuski, A. Shapiro, R. Gowrisankar, Y. Xu, T. Richmond, A. Feng, Mobile Personal Healthcare Mediated by Virtual Humans, Proceedings of the companion publication of the 19th international conference on Intelligent User Interfaces, Haifa, Israel, February 2014 paper, bibtex
 
  E. Miguel, A. Feng, Y. Xu, A. Shapiro, Towards Cloth-Manipulating Characters, ACM SIGGRAPH Conference on Motion in Games, Dublin, Ireland, November 2013
 
  Y. Xu, A. Feng, A. Shapiro, A Simple Method for High Quality Lip Syncing, Symposium of Interactive 3D Graphics and Games 2013, Orlando, Florida, March, 2013
 
  A. Shapiro, D. Chu, P. Faloutsos The Controller Toolkit, Symposium of Computer Animation 2006, Posters & Demos, Vienna, Austria, August 2006
 
  M. Kallmann, A. Shapiro, P. Faloutsos, Planning Motions in Motion, Symposium of Computer Animation 2006, Posters & Demos, Vienna, Austria, August 2006
 
  A. Shapiro, P. Faloutsos, Steps Toward Intelligent Interactive Control, Symposium of Computer Animation 2005, Posters & Demos, Los Angeles, CA, July 2005
 
  A. Shapiro, P. Faloutsos, Victor Ng-Thow-Hing, Dynamic Animation and Control Environment, Eurographics Symposium on Computer Animation, Posters & Demos, Grenoble, France, August 2004
 
  A. Shapiro, Y. Cao, P. Faloutsos, Stylistic Motion Decomposition, Eurographics Symposium on Computer Animation, Posters & Demos, Grenoble, France, August 2004


Book Chapters & Demos
  L.P. Morency, A. Shapiro, S. Marsella, Modeling Human Communication Dynamics for Virtual Humans, Coverbal Synchrony in Human-Machine Interaction, CRC Press, 2013
 


Credits
 
 Alvin and the Chipmunks: The Squeakquel, Rhythm & Hues Studios 2009 (feature film)
 The Incredible Hulk, Rhythm & Hues Studios 2008 (feature film)
 The Force Unleashed, Industrial Light & Magic/LucasArts 2008 (video game)


Media Coverage
 
 A Digital Media World magazine interview about my SIGGRAPH work in their December, 2003 edition.
 Cover graphics, UCLA Computer Science Department Handbook 2006 (pdf)


Past Advisor
 
 Professor Petros Faloutsos
  
Projects...
FreeDip
 A web-based version of the game of Diplomacy. Originally developed by Guy Tsafnat and myself, this version is written in Java and plugs into a JSP-compliant webserver. This was used as a testbed for my automated player and can currently self-play approximately 1000 games/day.
  
Academic Past
UCLA, 2001 - 2007
 Ph.D., Computer Science 2007
 M.S., Computer Science 2002
 
San Jose State University 1999 - 2001
 M.S. program in Computer Science
 Graduation with Guy Tsafnat, wearing the Master's hood.
 
University of California, Santa Cruz 1994 - 1996
 B.A. in Computer Science
 
University of Massachusetts, Amherst 1989 - 1991
  
  
Some Fun Memories
Maui 2000
Whitewater Rafting
 Sacramento 2001 for my friend's bachelor party. This part of the trip was called Chunder. Not all of us made it through the falls. Here's the entire sequence if you'd like to see it.
 
News
April 17, 2014 News article from USC's Viterbi School of Engineering on the Fast Avatar Capture and Simulation work: The full article is here.

February, 2014
Here is a Gizmodo article on our Rapid Avatar Capture and Simulation project where we can capture a person using a first-generation Microsoft Kinect system and simulate them in a matter of minutes.

GIZMODO - Kinect Avatar Creation at USC from Gizmodo on Vimeo.


We believe that this kind of capability dramatically changes the economics of avatar capture (essentially, it's now free and takes very little time) and will have an impact on 3D character acquisition and use going forward.
Here's a video of the entire capture process:


March, 2013
I volunteered for a project out of ICT's Graphics Lab. My face and performance were captured by their Light Stage (I was not involved with the technical aspects). Here are the results, as shown on different technological platforms by both Nvidia:


and Activision:



In the days following the capture, I would walk by my colleagues in the Graphics Lab (my office is very close to theirs) and they would be studying me very closely as I walked by. Occassionally they would say things like "You should see what we are doing to 'Ari' today.". In didn't take long before I insisted that they call my digital doppleganger 'Ira' instead of 'Ari' to loosen some of this association between myself and this digital version of myself. You can do what you want with Ira; it gets a little personal when you are doing it to 'me'. I'm sure that as this phenomena of capturing a person and digitizing them, then putting their digital version in various situations will lead to a number of psychological studies, particularly now that the distinction between the two is getting smaller and smaller.

In case you are curious, the 'yogurt parfait' incident came when the director (Oleg Alexander from ICT) asked me to get mad about something so that they could record some kind of emotional expression. About a week before the capture session, I had stopped by McDonalds in the morning for their $1 sausage muffin (substituting sausage for egg) and the $1 fruit and yogurt parfaits, as I had done so a few times a week for the past month. Usually the strawberries are a bit too cold, and sometimes frozen, so I would typically eat the yogurt, and would sometimes not even touch the strawberries, depending on how icy and cold they were. That one day, they gave me an entire plastic cup full of frozen, hard strawberries without a bit of yogurt, which I didn't realize until I left the drive through. I came back the next day, I asked for a refund, then asked the cashier to check the parfaits and make sure that there is enough yogurt in them. This turned into an unpleasant exchange with the manager on duty, who insisted that all parfaits are exactly the same, and that it would have been impossible to get a parfait that lacked yogurt, and refused to check any of the existing parfaits for their yogurt content (I still wanted another one...) I then wrote a complaint to McDonalds via email. They sent me a coupon for a free meal, told me they took my complaint seriously, and told me they would talk to the manager at that restaurant. That was about as much effort I wanted to put into a defective $1 purchase. I went back to that McDonalds several weeks later, and noticed the manager wearing a different, what appeared to be, a more formal, uniform, and the cashier also for the first time refused to substitute sausage for egg anymore in the $1 sausage mcmuffin. So I assume that someone talked to the owner and the manager, and among other things, a decision was made to not allow substitutions anymore. Not sure what happened to the parfaits - I stopped buying them. I suspect that my complaint set in motion a number of things. All in all, I stopped frequenting there for breakfast. So it's nice that Digital Ira can carry on my message without any additional effort on my part (how long do things last on the Internet, these days? Forever?), and stand up for the little guy against the corporate multinational.

Here's a fun parody of the yogurt ordeal:



Thoughts about the Ph.D. process Before I attended UCLA, I had no idea how to distinguish one university from another, not withstanding their rank in US News & World Reports. I sifted through as many graduate student web pages I could in an attempt to gain insight into the university, the program and its students. I found those student web pages that introduce their readers to the university and Ph.D. lifestyle were the most helpful during my search and choice of Ph.D. programs. With that in mind, I will do my best to return the favor and assume that many of you reading my web page are interested in UCLA's graduate program and are not attending this university.

Please keep in mind that this document is written from the perspective of a first second third fourth fifth sixth year Ph.D. student recent Ph.D. graduate someone who is several years outside of his Ph.D. My thoughts, opinions and conclusions will change as I learn, experience and continue to reflect on my tenure in a computer science Ph.D. program.

The Location. UCLA is located in a nice part of Los Angeles - Westwood. Westwood has many coffee shops and movie theatres. Indeed, many movie premieres are held here, red carpet and all. Not entirely a college town, there is a large commercial area in Westwood along the 'Wilshire Corridor' - a stretch of road that holds many high-rises and commercial buildings. Westwood is only 5 miles inland and the beach is a short ride away. Westwood is also adjacent to some of the well-to-do areas of Los Angeles: Bel Air, Beverly Hills and Brentwood. UCLA's local college rival is USC (University of Southern California), a private university located in downtown Los Angeles. They have a better football team and we have a better basketball team. As a private university with a large endownment, they have more money than we do, but we have a much better campus and the tuition at UCLA is a fraction of that at USC.

Los Angeles, on the whole, is an amazing city. Culturally, there is a lot to do here. The city is ripe with comedy clubs, playhouses, museums, exhibits and sporting events. There is a stigma attached to Los Angeles by those who don't live here that LA is smoggy, crowded and that the people are superficial.

Well, some of that is true.

However, the smog isn't bad in areas near the coast where UCLA resides, although it did take me about a year to get used to the change in air quality. The freeways and roads are frequently clogged, but after you live here for awhile you figure out faster ways to travel, usually on side roads. I have met some wonderful and interesting people in LA, including my beautiful wife.

Most people in the Greater Los Angeles live in high-density housing, such apartment buildings and condominiums. There are, however, many suburbs in Los Angeles that are not crowded and where parking is plentiful. Public transportation is fair and a student could arrange their living so that they take the bus everywhere they need to go. But since Los Angeles is so spread out, it is desirable to have access to a car. The advantage of high-density living is that you have restaurants, grocery stores and movie theatres all within walking distances of housing areas. Lastly, yes, Los Angeles is exceptionally image conscious. This is due in part because of the weather (it's nice and sunny, almost all the time) and partly due to the proximity to the television and movie industries.

However, because of the tremendous mix of different people, no one social group dominates the area. You meet a lot of people who do all sorts of different things for a living - doctors, lawyers, poets, journalists, actors and confused graduate students. LA is also a good place to go if you like to eat. There are thousands of restaurants from all different cultures (Chinese, Japanese, Vietnamese, Persian, Italian, French, Mexican, Thai - you name it). UCLA sits in an educated, urban area of Los Angeles. The congressional district that UCLA occupies even has the highest percentage of college graduates in California (at 44%). It also holds the second largest Jewish community in the United States (outside of New York).

There are many professional technology job opportunities available outside of the university. There are biotech, entertainment and financial companies that need help from computer scientists and engineers. Los Angeles is not, however, the hub of technology opportunities. The Silicon Valley is still a better place to find technology companies.

Teaching. UCLA is a public university and the undergraduate classes are large. This is an advantage for the graduate student who is interested in teaching. There are plenty of opportunities to become a teaching assistant (TA). Every UCLA Ph.D. student is required to become a TA for at least one quarter during their graduate tenure. However, being a TA is not as highly regarded as being a research assistant (RA). The primary motivation of a research university is research, not teaching. The right research position would allow you both to get paid and to further your research, which in turn will get you closer to finishing your Ph.D. A TA position will not do this for you. A TA position will require many hours of instruction, paper grading and preparation for the course you are assisting. It is not unusual to be a TA for a class that is not directly related to your research. Sometimes, however, your advisor will use the TA position as a way to indirectly fund their students. This is done by using the TA to lecture the classes while using graders (hired graduates, usually at about $10/hour or so) to perform test and homework grading. This means that the TA will need only to lecture once/week and hold office hours, thus reducing the workload significantly.

Some Statistics. Over the years I have seen UCLA rank somewhere above 10th and less than 15th for graduate programs in Computer Science. Reflecting on this, I think there are the 'top' institutions (MIT, Stanford, Berkeley, CMU, let's call them '1st tier') in the field, and then there are a bunch of other institutions ranked slightly below them (let's call them '2nd tier'), but it would be difficult to establish any preferred position of one over the other, since it depends greatly on the specific field, and frankly, the specific professor in that field. So whether you go to the university ranked number 8 or 18, it really doesn't matter much. There are many other tiers, but UCLA is way up there, and has a great reputation both locally, nationally and internationally.

If you're interested in admission statistics for UCLA, the following is data from 2007-2011 for the Computer Science department (29% admission rate). Note that the average time to a Doctoral Degree is 6 years, so get comfortable...
For undergaduates, UCLA was ranked as the second best public university in the nation (US News & World Report 2013). As an undergraduate institution, it is just as hard (or harder...)to get into than California's other top public university, Berkeley. See the stats. The admission rate for Californians at UCLA was 17.7% versus 22.7% at Berkeley, with an average GPA of 4.2.

Why did I choose UCLA? Since I wasn't sure which area of research I was most interested in (as most entering students aren't), I wanted to attend a university that was large enough to provide several different research areas. Smaller departments seemed too limiting, since each research area in those universities had only one or two professors. This meant that if you didn't like/get along with/were interested in one of the faculty, then you were stuck either choosing an area that you aren't passionate about, or you spent time working in an area outside of the expertise of your university, which is difficult to do. This is not a problem that can be completed avoided when choosing to attend a large university. It is not unusual at all to begin work in one research area, then switch to another one that you find to be more interesting. However, transferring to another graduate school has its own cost; each graduate school has their own set of requirements that often do not overlap with your own school, so students that transfer end up taking many additional tests and classes. Transfers from other University of California schools have an advantage at UCLA: many of their classes taken as graduates are transferable to UCLA's program.

UCLA's location was another influencing factor. As a professional with many years of experience, I wanted to live in an area that could support a technology industry that I could either consult with or work for during the summers. If you attend a university in a remote area, you will need to leave that area during the summer if you don't have employment with the university itself. For example, I spoke with to a number of students at UCSB in 2001 (observation 8/11/05: wow, I need to update this web page - this info is 5 years old! observation 3/30/13: looks like I haven't updated this information in 12 years...) who all went to the Silicon Valley for their summer internships - a 5 hour drive away. In Los Angeles, there is the Jet Propulsion Laboratory (JPL) in Pasadena and the Hughes Research Center (HRC) in Malibu which employ students for internships and allow people to stay in the area. Also, since UCLA is a public university and I am a California resident, it might be possible to work outside of school and still afford it's meager tuition rates.

In addition, UCLA is close to many other research universities. The University of Southern California (USC) is only a 20 minute drive away and it is common for students to attend talks there. Caltech is less than an hour away, UCSD is three hours away and UCSB is two hours away. Plus, Northern California is less than an hour away by plane and holds other top-notch univerities such as Stanford and Berkeley.

Although reputation is an important consideration when choosing a school, you need to also consider where you are going to live for the next four or five years of your life. (The temperature during the winter in LA is about 60 degrees and you never have to shovel any snow!). I value urban areas that are culturally diverse.

Attrition It is difficult to collect statistics on graduate school attrition rates (how many people leave the program and never finish). I would estimate that approximately 1/3 to 1/2 of those who enter the Ph.D. program do not finish. Some leave with master's degrees, some get industry jobs and never return, others simply don't progress in their research and quietly drop out. I don't believe that the attrition rate at UCLA is any different than other top universities. I've spoken to others at different universities, who say that 50% attrition sounds about right. Keep this in mind, you who is thinking about pursuing a PhD - it's a marathon.

The secret to affording graduate school:Unlike business school, medical school or law school, graduate students in the sciences get paid to go to attend the university. They are paid through a number of different sources: teaching assistantships (teaching classes), research assistantships (work for a professor in their area of research) or fellowships (free money). Although universities do charge tuition, these funding sources include a tution waiver. The pay is nominal, often less than $25k/year, but is enough money to live on for a short period of time. Of course, $25k goes a lot further in some places than in others. It will go a lot further in a rural area than an urban one. Ph.D. students are typically given funding priority and not all Master's students will get tution waivers through academic employment. This is particularly true of out-of-state and foreign students, since the university must pay an additional fee for out-of-state-tuition for them. California public universities are constantly fighting budget cuts, and priority is given to in-state students.

Even though we scientists get paid to go to school, we aren't rewarded upon graduation as our colleagues who go to professional training schools do. Lawyers and MBA's graduating from the equivalent level of program in law and business as UCLA is in computer science will receive a six figure salary as a starting salary and will receive even higher salaries as their career progresses.


Papers, Symposiums, Invited Talks, Presentations and Journals Every academic has a publication list that tells you about the kind of research that they have done and the major areas of interest for that research. These publications are obtained by 1) doing research, 2) sending that work into a conference/journal where it is reviewed by others in the field who are very knowledgable about the particular topic of research. The reviewers typically rate the work, make specific comments about the content and express whether or not they feel the work is of publication quality for that particular conference/journal. Publications that are accepted typically have to be presented at the conference where there is an opportunity to present the idea as well as field questions about the work.

Each field has its own conferences and journals that are seen a more or less prestigious than others. Similarly, each conference/symposium/journal has a different acceptance rate, ranging from very high (almost everything sent is accepted) to very low (say, only 10% of submissions are accepted). Conferences often have a "best paper" award, which is judged as the best paper of those accepted at the conference.

The submission process for journal publications generally takes longer than the conference submission process. In addition, journal papers are not accompanied by a live presentation. Unlike conferences, journals do not have strict deadlines. The reviewing process is more involved, with many changes to the submission taking place in order to satisfy the reviewer's criticisms. As such, journals take longer to publish than do conferences, and the information can often be revealed at a conference much earlier than by the time the journal is published and distributed. Computer Science research advances fairly quickly, so new work typically comes out in conferences rather than journals. Journals are often used as a means of archiving the ideas contained in the conference papers, containing greater details of those works.

There is no strict requirement for publishing in order to finish your Ph.D., except for your dissertation. Student wishing to pursue an academic career should develop a good publication record so that they become known within their field for their work. In addition, conferences are great opportunities to meet your peers in the field, socialize, make friends and develop collaborative opportunities.