סמינר מחלקתי

18 בדצמבר 2014, 13:00 
חדר 206 בניין וולפסון  
סמינר מחלקתי

Machine-Learning the Hidden Universal Semantics of Natural Languages

Dr.  Omri Abend - Edinburgh University

Abstract:

 

 The field of Natural Language Processing (NLP) has recently been pivotal in producing important language technologies such as machine translation and question answering. Such technologies are based on elaborate structural representations of text, detected by  statistical methods. However, common approaches to structural representation are language-specific or even domain-specific, limiting the applicability of NLP tools and models. How to represent both the idiosyncrasies of specific domains and languages as well as their commonalities is still an open question.

 

 In my talk I will address these questions and propose an approach for discovering a level of representation shared by all languages using latent variable models for structured prediction. Under this approach, learning starts from universally-applicable coarse-grained logical structure, which is used to bootstrap the learning of more fine-grained semantic distinctions, as well as the learning of the specifics of individual languages. I will discuss the value of universal semantic structures both to the computational modeling of child language acquisition, and to leading NLP applications, focusing on machine-reading of web data and machine translation.

 

 

 Joint work with Ari Rappoport, Shay Cohen and Mark Steedman.

 

 Short Bio:

 Omri Abend is a postdoctoral researcher in the School of Informatics of the University of Edinburgh, working in the fields of Natural Language Processing and Computational Linguistics in Mark Steedman's lab. Omri earned his PhD from the Hebrew University, under the supervision of Ari Rappoport. During his PhD studies, he was a member of the Azrieli Fellows Program for promoting academic excellence and leadership. Prior to that, Omri studied mathematics and cognitive sciences at the Hebrew University (BSc, summa cum laude).

 

 

ההרצאה תתקיים ביום חמישי, 18.12.14, בשעה 13:00 בחדר 206, בנין וולפסון הנדסה, הפקולטה להנדסה, אוניברסיטת תל-אביב.

 

1.1.15

01 בינואר 2015, 15:00 
Kitot 011  
1.1.15

You are invited to attend a lecture

By

Dr. Alex Ruderman, Associate Professor

Department of Electrical and Electronic Engineering, School of Engineering, Nazarbayev University Kabanbay Batyr Ave 53 Astana, Kazakhstan 010000.

Phone +7-7172-709146

Email alexander.ruderman@nu.edu.kz

 

 

On the subject:

 

Time Domain Analysis of Multilevel Converter Voltage and Current Quality

Today power electronics community is biased towards frequency domain analysis of PWM phenomena. The seminar goal is to show that time domain averaging methods are more adequate for certain integral problems that don't require a specific knowledge of individual harmonics. The power of time domain analysis is demonstrated while elaborating voltage and current quality in multilevel inverters for PWM and staircase (step) modulation. For relatively high frequency PWM, the solution is obtained under the asymptotic assumption – a switching frequency is supposed much higher than a fundamental one. The closed-form asymptotic solution for an arbitrary voltage level count involves only elementary functions and, in fact, presents voltage quality upper bound for nearest level PWM. The asymptotic solution is practically very accurate for relatively high switching-to-fundamental frequency ratios (>25-30). For staircase modulation, voltage and current optimal quality (minimal Total Harmonic Distortion - THD) problems are formulated in time domain as constrained optimization ones. Optimal switching angles that minimize voltage and current THD accounting for all switching harmonics are found using numerical optimization for different inverter level counts and the whole modulation index range 0<M<1. Suggested are simple voltage and current THD ripple-free approximations for arbitrary level counts and modulation indices. While these simple hyperbolic formulas accurately approximate the average trend, local maxima and minima of voltage and current THD may be addressed on separate. The obtained current THD is actually frequency weighted THD that assumes inductance dominated RL-load. For grid-connected applications, normalized current fundamental harmonic is smaller than voltage modulation index and current THD will be larger than frequency weighted voltage one.

 

 

1 January 2014, at 15:00,

Room 011, Kitot Building

 

ברכות לזוכים וותודה מקרב לב על העבודה הקשה ועל ההצטיינות.

14 דצמבר 2014

מורים מצטיינים בהוראה – תשע"ד

 

מצטייני רקטור  בהוראה:

סגל בכיר

סגל זוטר

פרופ' אבישי אייל – בית הספר להנדסת חשמל

ד"ר עופר עמרני – בית הספר להנדסת חשמל

פרופ' אברהם קריבוס – בית הספר להנדסה מכנית

אוהד איזנהנדלר – המחלקה להנדסת תעשייה

גב' שני ארדזי – בית הספר להנדסת חשמל

גב' מעיין מלכי – בית הספר להנדסה מכנית

מצטייני דקאן  בהוראה:
ביה"ס להנדסת חשמל
סגל בכיר

סגל זוטר

מורה מצטיין:

ד"ר ערן סוחר

 

ציון לשבח:

פרופ' אריה ירדאור

פרופ' משה טור

פרופ' יוסי שחם

מתרגל מצטיין:

רועי שילה

 

ציון לשבח:

ניצן הבלר

אופיר וייס

עמית לוי

דנה מנקר

עומר קוטליצקי

אסיה שפירא

   

ביה"ס להנדסה מכנית

סגל בכיר

סגל זוטר

מורה מצטיין:

ד"ר יאיר שוקף

 

 

ציון לשבח:

 פרופ' ראמי חג- עלי

 פרופ' ויאצסלב קרילוב 

מתרגל מצטיין:

גב' רמה סנדובסקי

 

ציון לשבח:

מר יונתן הורוביץ

מר ערן שחורי

   

המחלקה להנדסת תעשיה

סגל בכיר

סגל זוטר:

מורה מצטיין:

ד"ר ערן טוך

 

ציון לשבח:

גב' רעות נחם  

מתרגל מצטיין:

מר יובל אלבר

 

ציון לשבח:   

מר עומר דמבינסקי

 
המחלקה להנדסה ביו רפואית
סגל בכיר סגל זוטר:

מורה מצטיין:

ד"ר שרון זלוצ'יבר

 

ציון לשבח:

ד"ר רונן שושניק

 

מתרגל מצטיין:

מר דוד מזרחי

 

ציון לשבח:

גב' רננה סבי

קורסי שירות ומורים מן החוץ

מורים מצטיינים:

מתרגלים מצטיינים:

ד"ר גלית אשכנזי גולן -  החוג לסטטיסטיקה

פרופ' אשל בן-יעקוב   - ביה"ס לפיזיקה ואסטרונומיה

ד"ר אלי להר             - ביה"ס למתמטיקה

גב' אלה מרקמן זמיר  - ביה"ס להנדסה מכנית

מר יותם סמילנסקי     - ביה"ס למתמטיקה

מר רן עזורי              - ביה"ס למתמטיקה

גב' עדי אשכנזי  - ביה"ס לפיזיקה ואסטרונומיה

גב' ענת אמיר    - ביה"ס למתמטיקה

גב' אור ברוך     - ביה"ס למתמטיקה

מר אופיר הררי  - החוג לסטטיסטיקה

מר דויד לגזיאל  - החוג לסטטיסטיקה

מר אורן סלון     - ביה"ס לפיזיקה ואסטרונומיה

 

Seminar 25.12.14

25 בדצמבר 2014, 15:00 
Kitot 011  
Seminar 25.12.14
 

 

You are invited to attend a lecture

By

 

Dr. Ido Kaminer, MIT

 

On the subject:

 

Quantum Theory of Čerenkov Radiation, 

Spectral Cutoff and Resonances

When a charged particle travels faster than the phase velocity of light in a medium, it produces Čerenkov radiation. We show that the Čerenkov Effect contains new phenomena arising from the quantum nature of a charged particle. Specifically, with proper design of particle wavepacket, we predict the traditional Čerenkov radiation angle splits into two distinctive cones of photonic shockwaves. One of the shockwaves can move along a backward cone, otherwise considered impossible for Čerenkov radiation in ordinary matter. More importantly, and more generally, the spectral response reveals an upper frequency cutoff at which the photon emission rate is diverging – manifesting a new resonant light-matter interaction. The origins of the major deviations from the known understanding of the Čerenkov effect are the nonlinear nature of the interaction, and its spatial extent that exceeds the wavelength of the emitted radiation. Importantly, our findings are observable for electron beams with realistic parameters, offering new applications including coherent x-ray sources and open a new realm for Čerenkov detectors.

 

 

25 December 2014, at 15:00,

Room 011, Kitot Building

 

Seminar 18.12.14

18 בדצמבר 2014, 15:00 
Kitot 011  
Seminar 18.12.14

  Physical Electronics Dept.

 

You are invited to attend a lecture

By

 

Ofer Levi

Associate Professor, The Department of Electrical and Computer Engineering and The Institute of Biomaterials and Biomedical Engineering, The University of Toronto ofer.levi@utoronto.ca

 

On the subject:

Portable optical sensors and imaging systems for biomedical applications

 

Optical techniques are widely used in clinical settings and in biomedical research to interrogate bio-molecular interactions and to evaluate tissue dynamics. Miniature integrated optical systems for sensing and imaging can be portable, enabling long-term studies in living tissues.

We present the development of a compact multi-modality optical neural imaging system, to image tissue blood flow velocity and oxygenation, using a fast CCD camera and miniature VCSEL illumination. We combined two techniques of laser speckle contrast imaging (LSCI) and intrinsic optical signal imaging (IOSI) simultaneously, using these compact laser sources, to monitor induced cortical ischemia in a full field format with high temporal acquisition rates. We have demonstrated tracking seizure activity, evaluating blood-brain barrier breaching, and integrating fast spatial light modulators for extended imaging depth and auto-focusing during brain imaging of flow dynamics. Our current studies include prototype designs and system optimization and evaluation for a low-cost portable imaging system as a minimally invasive method for long-term neurological studies in un-anesthetized animals. This system will provide a better understanding of the progression and treatment efficacy of various neurological disorders, in freely behaving animals. Finally, we will review our progress in optical nano-structure sensor fabrication and optimization for bio-sensing of molecular interactions.

Biography

Dr. Ofer Levi is an Associate Professor in the Institute of Biomaterials and Biomedical Engineering and the Edward S. Rogers Sr. Department of Electrical and Computer Engineering at the University of Toronto. Dr. Levi received his Ph.D. in Physics from the Hebrew University of Jerusalem, Israel in 2000, and worked in 2000-2007 as a Postdoctoral Fellow and as a Research Associate at the Departments of Applied Physics and Electrical Engineering, Stanford University, CA.  He serves as an Associate Editor in Biomedical Optics Express (OSA) and is a member of OSA, IEEE-Photonics, and SPIE.  His recent research areas include biomedical imaging systems and optical bio-sensors based on semiconductor devices and nano-structures, and their application to bio-medical diagnostics, in vivo imaging, and study of bio-molecular interactions.  More details can be found at http://biophotonics.utoronto.ca/

 

 

Thursday, December 18, 2014 at 15:00, Room 011, Kitot Eng. bldg.

 

 

סמינר מחלקתי Prof. Brad Nelson

14 בינואר 2015, 15:00 
וולפסון 206  
0
סמינר מחלקתי Prof. Brad Nelson

 

 

 

 

School of Mechanical Engineering Seminar
Wednesday, January 14, 2014 at 15:00
Wolfson Building of Mechanical Engineering, Room 206

 

 

 

MicroRobotics and NanoMedicine

 

Prof. Brad Nelson

Institute of Robotics and Intelligent Systems

ETH Zurich

Switzerland

 

 

While the futuristic vision of micro and nanorobotics is of intelligent machines that navigate throughout our bodies searching for and destroying disease, we have a long way to go to get there. Progress is being made, though, and the past decade has seen impressive advances in the fabrication, powering, and control of tiny motile devices. Much of our work focuses on creating systems for controlling micro and nanorobots as well as pursuing applications of these devices. As systems such as these enter clinical trials, and as commercial applications of this new technology are realized, radically new therapies and uses will result that have yet to be envisioned.

 

 

About the speaker:

Brad Nelson received mechanical engineering degrees from the University of Illinois (B.S. 1984) and the University of Minnesota (M.S. 1987), and a Ph.D. in Robotics (School of Computer Science) from Carnegie Mellon University (1995). He has been the Professor of Robotics and Intelligent Systems at ETH Zürich since 2002 and has received a number of awards for his work in robotics, nanotechnology, and biomedicine. He serves on the advisory boards of a number of academic departments and research institutes across North America, Europe, and Asia and is on the editorial boards of several academic journals.

Prof. Nelson has been the Department Head of Mechanical and Process Engineering at ETH, Chairman of the ETH Electron Microscopy Center, and is a member of the Research Council of the Swiss National Science Foundation. He is a member of the board of directors of three Swiss companies.

Before moving to Europe, Prof. Nelson worked as an engineer at Honeywell and Motorola and served as a United States Peace Corps Volunteer in Botswana, Africa. He has also been a professor at the University of Minnesota and the University of Illinois at Chicago.

 

סמינר מחלקתי

16 בדצמבר 2014, 14:00 
חדר 206 בניין וולפסון  

Intelligent internally-illuminated pavement markers for lane delineation

 

ד"ר עמית שחר - מעבדת תפעול, תפיסה, סימולטורים וסימולציה (LEPSIS)

המכון הצרפתי למדע וטכנולוגיה של תחבורה, תכנון ורשתות   (IFSTTAR)

 

Abstract:

 

Road delineation treatments provide drivers with visual cues, which can enhance vehicle navigation and control. However, they can lead to undesirable effects and increase risk, due to behavioural adaptation or if not properly designed or installed. In recent years, the use of internally-illuminated pavement markers for road delineation has become increasingly popular. INtelligent Renewable Optical ADvisory System (INROADS) is a research and development cooperation project carried out within the 7th Framework Programme of the European Commission, dedicated to the development of intelligent road lighting applications. Within that framework, an Active Lane Delineation (ALD) application is being developed: internally-illuminated pavement markers are automatically turned on to outline the lane and road edges as a vehicle approaches to and passes through curves, and are turned off otherwise. In this talk, following an introduction on INROADS, delineation treatments and country road crashes, I will present the ALD prototype designed at The French Institute of Sciences and Technology for Transport, Development and Networks (IFSTTAR) for INROADS and a simulator experiment designed (at IFSTTAR) to shed some light on its relative effectiveness. The experiment compared night-time driving on a country road in the presence of the ALD application to driving on an unlit road and to a road illuminated by conventional road luminaries, on a number of objective and subjective criteria. At a subjective level, the participants perceived driving in the unlit condition, as less safe, less comfortable and allowing less control than in both the ALD and luminaries conditions, but there were no significant differences between these latter conditions. However, all in all the behavioral results demonstrated safer performance measures of vehicle trajectory, speed and control, in the ALD condition relative to both the unlit condition and the luminaries condition. Human factors considerations on the design of internally-illuminated pavement markers for road delineation will be discussed, focusing on the cost-benefit trade-off of the illumination effects on driver perception and driving behavior.

 

 

 

ההרצאה תתקיים ביום שלישי, 16.12.14, בשעה 14:00 בחדר 206, בנין וולפסון הנדסה, הפקולטה להנדסה, אוניברסיטת תל-אביב.

EE Seminar: Dr. Itai Dinur

~~
(The talk will be given in English)

Speaker: Dr. Itai Dinur,
Department of Computer Science, Ecole Normale Superieure, Paris, France

Monday, January 5th, 2015
15:00 - 16:00
Room 011, Kitot Bldg., Faculty of Engineering

Thinking out of the Encryption Box
Abstract
Block ciphers are the main tool used today to protect sensitive data. They are typically constructed by combining several simpler components in order to provide the highest possible level of security. In this talk I will consider these components as black boxes, and analyze the security of the overall structure while ignoring the low-level details of each black box.
More specifically, I will describe new attacks that break three classical block cipher constructions more efficiently than all the previously found attacks. Interestingly, some of the new techniques can also be used to solve certain well known algorithmic problems which have nothing to do with cryptography.

The talk will require no prior knowledge in cryptography.

05 בינואר 2015, 15:00 
חדר 011, בניין כיתות-חשמל  

סמינר מחלקתי Rahel Ofir

07 בינואר 2015, 15:00 
וולפסון 206  
0
סמינר מחלקתי Rahel Ofir

EE Seminar: Dr. Eitan Yaakobi,, Technion

~~(The talk will be given in English)

Speaker: Dr. Eitan Yaakobi,
Computer Science Department, Technion

Monday, December 15th, 2014
15:00 - 16:00
Room 011, Kitot Bldg., Faculty of Engineering

Generalized Sphere Packing Bound

Abstract
Kulkarni and Kiyavash recently introduced a new method to establish upper bounds on the size of deletion-correcting codes. This method is based upon tools from hypergraph theory. The deletion channel is represented by a hypergraph whose edges are the deletion balls (or spheres), so that a deletion-correcting code becomes a matching in this hypergraph. Consequently, a bound on the size of such a code can be obtained from bounds on the matching number of a hypergraph. Classical results in hypergraph theory are then invoked to compute an upper bound on the matching number as a solution to a linear-programming problem: the problem of finding fractional transversals.

The method by Kulkarni and Kiyavash can be applied not only for the deletion channel but also for other error channels. This work studies this method in its most general setup. First, it is shown that if the error channel is regular and symmetric then the upper bound by this method coincides with the well-known sphere packing bound and thus is called the generalized sphere packing bound. Even though this bound is explicitly given by a linear programming problem, finding its exact value may still be a challenging task. The art of finding the exact upper bound (or slightly weaker ones) is the assignment of weights to the hypergraph’s vertices in a way that they satisfy the constraints in the linear programming problem. In order to simplify the complexity of the linear programming, we present a technique based upon graph automorphisms that in many cases significantly reduces the number of variables and constraints in the problem. We then apply this method on specific examples of error channels. We start with the Z channel and show how to exactly find the generalized sphere packing bound for this setup. Next studied is the non-binary limited magnitude channel both for symmetric and asymmetric errors, where we focus on the single-error case. We follow up on the deletion channel, which was the original motivation of the work by Kulkarni and Kiyavash, and show how to improve upon their upper bounds for single-deletion-correcting codes. Since the deletion and grain-error channels resemble a very similar structure for a single error, we also improve upon the existing upper bounds on single-grain error-correcting codes. Finally, we apply this method for projective spaces and find its generalized sphere packing bound for the single-error case.

This is a joint work with Arman Fazeli and Alexander Vardy from the University of California, San Diego

15 בדצמבר 2014, 15:00 
חדר 011, בניין כיתות חשמל  

עמודים

אוניברסיטת תל אביב עושה כל מאמץ לכבד זכויות יוצרים. אם בבעלותך זכויות יוצרים בתכנים שנמצאים פה ו/או השימוש שנעשה בתכנים אלה לדעתך מפר זכויות
שנעשה בתכנים אלה לדעתך מפר זכויות נא לפנות בהקדם לכתובת שכאן >>