Date (
Fall 2011
) 
Category 
Seminar Info 
2011/12/05 
Vision, Graphics, and Robotics 
Place: ENGMC 437
Time: 11:00  12:00
Speaker:
Remi
Ronfard
Affiliation: INRIA RhoneAlpes
Area:
Graphics
Title: Computational Model of Film Editing for Interactive Storytelling
Abstract: > In interactive storytelling, it is useful to present 3D animation in a cinematic style, which means selecting appropriate cameras and appropriate intercutting between cameras to properly convey the narrative. We propose an optimization framework for selecting shots and cuts while the narrative unfolds, based on a relatively simple scoring scheme driven by working practices of film and television. We cast the problem of film editing as selecting a path in time through a collection of evolving film takes (a take is a continuous sequence of images from a given camera) and precisely deciding when to cut in and out of film takes. In contrast to related work, we also account for a precise enforcement of pacing (rhythm at which cuts are performed). We propose an algorithm suitable for online editing which uses an efficient bestfirst search technique. The algorithm relies on shortterm anticipation to improve quality in cuts and produce movies consistent with the rules of cinematography and editing, including shot composition, continuity editing and pacing.

2011/10/28 
CQIL  Cryptography and Quantum Information 
Place: Rutherford Physics 112
Time: 15:30  16:30
Speaker:
John
Preskill
Affiliation: Caltech
Area:
Aisenstadt Chair Lecture
Title: Battling decoherence: the faulttolerant quantum computer
Abstract: Largescale quantum computers, if and when we succeed in building them,
will be able to solve problems that are beyond the reach of ordinary
digital computers. But constructing practical quantum computers will be
tremendously challenging. A particularly daunting difficulty is that
quantum computers are far more susceptible to making errors than
conventional digital computers. I will explain the principles of
quantum error correction and faulttolerant quantum computation, which
can enable a properly designed quantum computer with imperfect
components to achieve good reliability, and I will discuss the status of
current research on this topic.

2011/10/26 
CQIL  Cryptography and Quantum Information 
Place: Pavillon Andre Aisenstadt 1360, U de Montreal
Time: 16:00  17:00
Speaker:
John
Preskill
Affiliation: Caltech
Area:
Aisenstadt Chair Public Lecture
Title: Putting weirdness to work: quantum information science
Abstract: Quantum information (information encoded in a quantum system) has weird
properties that contrast sharply with the familiar properties of
classical information. Physicists, who have relished this weirdness for
many years, have recently begun to recognize that we can put the
weirdness to work  there are tasks involving the transmission and
processing of information that are achievable in principle because
Nature is quantum mechanical, but that would be impossible in a less
weird classical world. The security of cryptographic protocols that use
quantum information instead of classical bits can be founded on
principles of fundamental physics rather than assumptions about the
resources available to a potential adversary. A quantum computer, which
processes quantum information, could easily perform certain types of
calculations that would take far longer than the age of the universe on
today's supercomputers. However, constructing practical quantum
computers will be tremendously challenging.

2011/10/24 
CQIL  Cryptography and Quantum Information 
Place: Pavillon Andre Aisenstadt 6214, U de Montreal
Time: 16:00  17:00
Speaker:
John
Preskill
Affiliation: Caltech
Area:
Aisenstadt Chair Lecture
Title: Protected gates for superconducting qubits
Abstract: I explain how continuousvariable quantum errorcorrecting codes can be
invoked to protect quantum gates in superconducting circuits against
thermal and Hamiltonian noise. The gates are executed by turning on and
off a tunable Josephson coupling between an LC oscillator and a qubit or
pair of quits; assuming perfect qubits, we show that the gate errors are
exponentially small when the oscillator's impedance is large in natural
units. The protected gates are not computationally universal by
themselves, but a scheme for universal faulttolerant quantum
computation can be constructed by combining them with unprotected noisy
operations.

2011/10/20 
CQIL  Cryptography and Quantum Information 
Place: Pavillon AndreAisenstadt 6214, U de Montreal
Time: 16:00  17:00
Speaker:
Renato
Renner
Affiliation: ETH Zurich
Area:
Aisenstadt Chair Lecture
Title: An InformationTheoretic View on Thermalization
Abstract: How can the reversible dynamics of physical processes give rise to irreversible phenomena such as thermalization? I will reconsider this old question using modern tools from quantum information theory. In particular, I will show how recently developed techniques, such as the decoupling approach to quantum information, allow us to make quantitative statements about the thermalization properties of physical systems.

2011/10/12 
CQIL  Cryptography and Quantum Information 
Place: Pavillon AndreAisenstadt 1360, U de Montreal
Time: 16:30  17:30
Speaker:
Renato
Renner
Affiliation: ETH Zurich
Area:
Aisenstadt Chair Public Lecture
Title: What Does Quantum Cryptography Tell Us About Quantum Physics?
Abstract: Heisenberg's uncertainty principle asserts that certain observable quantities, such as position and velocity of a quantum particle, cannot be simultaneously known to arbitrary precision. More than 25 years ago, it has been realized that this feature of quantum physics can be exploited in cryptography for securely transmitting secret messages over insecure channels. Since this discovery, research in the emerging area of quantum cryptography has resulted in a variety of remarkable insights on the nature of information. In my talk, I will show how these insights influence our understanding of physics. In particular, using cryptographic considerations, I will argue that quantum physics is the most informative theory that is compatible with experimental observations and therefore, in a certain sense, complete. (This lecture does not require prior knowledge of quantum physics.)

2011/10/11 
CQIL  Cryptography and Quantum Information 
Place: McConnell 103
Time: 14:30  15:30
Speaker:
Nilanjana
Datta
Affiliation: Cambridge University
Title: Relative entropies and entanglement monotones
Abstract: We introduce two generalized relative entropies which have important
operational meanings in the context of state discrimination.
They also act as parent quantities for optimal oneshot rates of
various informationprocessing tasks in Quantum Information Theory.
Moreover, they lead us to define entanglement monotones which have interesting operational interpretations.

2011/10/05 
CQIL  Cryptography and Quantum Information 
Place: Pavillon AndreAisenstadt 6214, U de Montreal
Time: 16:00  17:00
Speaker:
Renato
Renner
Affiliation: ETH Zurich
Area:
Aisenstadt Chair Lecture
Title: Free Randomness Amplification
Abstract: Assume that we have access to a source of weakly random bits, with the only guarantee that the entropy of each bit (conditioned on all previously available information) is above a certain threshold. In 1984, Santha and Vazirani showed that, in a purely classical setting, it is impossible to amplify the quality of such a source. More precisely, there is no (deterministic) function that transforms weakly random bits into (almost) uniform ones. In this talk, I will consider a quantumphysical scenario and show that, using entanglement, the amplification of weakly random sources becomes possible. An interesting implication of this result is that no perfect randomness is required for applications such as cryptography or for Bell tests.

2011/09/27 
CQIL  Cryptography and Quantum Information 
Place: MC103
Time: 14:00  15:00
Speaker:
Mike
Duff
Affiliation: Imperial College
Title: Black holes and qubits
Abstract: Two different branches of theoretical physics, string theory and quantum
information theory (QIT), share many of the same features, allowing
knowledge on one side to provide new insights on the other. In
particular the matching of the classification of black holes and the
classification of four‐qubit entanglement provides a falsifiable
prediction of string theory in the field of QIT.
