2011-2012 Bulletin – PDF - SEAS Bulletin - Columbia University

118
range imaging, mosaicing and superresolution.
Classes are seminars with the instructor, guest
speakers, and students presenting papers and
discussing them.
COMS E6733x or y 3D photography
3 pts. Lect: 2.
Prerequisite: Experience with at least one of the
following topics: Computer graphics, computer
vision, pixel processing, robotics or computeraided
design, or permission of instructor.
Programming proficiency in C, C++, or JAVA.
3D Photography—the process of automatically
creating 3D, texture-mapped models of objects
in detail. Applications include robotics, medicine,
graphics, virtual reality, entertainment and digital
movies etc. Topics include 3D data acquisition
devices, 3D modeling systems and algorithms
to acquire, create, augment, manipulate, render,
animate and physically build such models.
COMS E6734y Computational photography
3 pts. Lect: 3.
Prerequisites: COMS W4160, W4731,
or a working knowledge of photography
are recommended. Students should have
knowledge in any of three core areas: computer
vision, computer graphics, or photography.
Computational techniques are used to produce a
new level of images and visual representations.
Topics include HDR imaging, feature matching
using RANSAC, image mosaics, image-based
rendering, motion magnification, camera lens
arrays, programmable lighting, face detection,
single and multiview geometry, and more.
COMS E6735y Visual databases
3 pts. Lect: 3. Not offered in 2011–2012.
Prerequisites: COMS W3133, W3134, or W3137
required. COMS W4731 and W4735 helpful but
not required. Contact instructor if uncertain.
The analysis and retrieval of large collections
of image and video data, with emphasis on
visual semantics, human psychology, and
user interfaces. Low-level processing: features
and similarity measures; shot detection; key
frame selection; machine learning methods
for classification. Middle-level processing:
organizational rules for videos, including
unedited (home, educational), semiedited
(sports, talk shows), edited (news, drama);
human memory limits; progressive refinement;
visualization techniques; incorporation of audio
and text. High-level processing: extraction of
thematic structures; ontologies, semantic filters,
and learning; personalization of summaries and
interfaces; detection of pacing and emotions.
Examples and demonstrations from commercial
and research systems throughout. Substantial
course project or term paper required.
COMS E6737x or y Biometrics
3 pts. Lect: 3.
Prerequisite: Background at the sophomore
level in computer science, engineering, or like
discipline. Corequisites: None In this course we
will explore the latest advances in biometrics as
well as the machine learning techniques behind
them. Students will learn how these technologies
work and how they are sometimes defeated.
Grading will be based on homework assignments
and a final project. There will be no midterm
or final exam. This course shares lectures with
COMS W4737. Students taking COMS E6737
are required to complete additional homework
problems and undertake a more rigorous final
project. Students will only be allowed to earn credit
for COMS W4737 or COMS E6737 but not both.
CSEE E6824y Parallel computer architecture
3 pts. Lect: 2.
Prerequisite: CSEE W4824. Parallel computer
principles, machine organization and design of
parallel systems including parallelism detection
methods, synchronization, data coherence and
interconnection networks. Performance analysis
and special purpose parallel machines.
CSEE E6831y Sequential logic circuits
3 pts. Lect: 3. Not offered in 2011–2012.
Prerequisite: CSEE W3827 or any introduction
to logic circuits. Generation and manipulation
of flow table descriptions to asynchronous
sequential functions. Coding of flow tables to
satisfy various design criteria. Delays, races,
hazards, metastability. Analysis of latches to
determine key parameters. Bounds of input
rates. Clocking schemes for synchronous
systems. Synthesis of self-timed systems using
4-phase or 2-phase handshakes.
CSEE E6832x or y Topics in logic design
theory
3 pts. Lect: 3.
Prerequisite: CSEE W3827 or any introduction to
logic circuits. A list of topics for each offering of
the course is available in the department office
one month before registration. May be taken
more than once if topics are different Iterative
logic circuits applied to pattern recognition. Finite
state machines; alternative representations,
information loss, linear circuits, structure theory.
Reliability and testability of digital systems.
CSEE E6847y Distributed embedded systems
3 pts. Lect: 2. Not offered in 2011–2012.
Prerequisite: Any COMS W411X, CSEE W48XX,
or ELEN E43XX course, or permission of
instructor. An interdisciplinary graduate-level
seminar on the design of distributed embedded
systems. System robustness in the presence
of highly variable communication delays and
heterogeneous component behaviors. The
study of the enabling technologies (VLSI
circuits, communication protocols, embedded
processors, RTOSs), models of computation,
and design methods. The analysis of modern
domain-specific applications including on-chip
micro-networks, multiprocessor systems, faulttolerant
architectures, and robust deployment
of embedded software. Research challenges
such as design complexity, reliability, scalability,
safety, and security. The course requires
substantial reading, class participation and a
research project.
CSEE E6861y Computer-aided design of
digital systems
3 pts. Lect: 2.
Prerequisites: (i) one semester of advanced
digital logic (CSEE W4823 or equivalent,
or instructor’s permission); and (ii) a basic
course in data structures and algorithms
COMS W3133, 3134, 3137, 3139, or 3157, or
equivalent, and familiarity with programming.
Introduction to modern digital CAD synthesis
and optimization techniques. Topics include:
modern digital system design (high-level
synthesis, register-transfer level modeling,
algorithmic state machines, optimal scheduling
algorithms, resource allocation and binding,
retiming), controller synthesis and optimization,
exact and heuristic two-level logic minimization,
advanced multilevel logic optimization, optimal
technology mapping to library cells (for delay,
power and area minimization), advanced
data structures (binary decision diagrams),
SAT solvers and their applications, static
timing analysis, and introduction to testability.
Includes hands-on small design projects using
and creating CAD tools.
EECS E6870x or y Speech recognition
3 pts. Lect: 3.
Prerequisites: Basic probability and statistics.
Theory and practice of contemporary
automatic speech recognition. Gaussian
mixture distributions, hidden Markov models,
pronunciation modeling, decision trees, finitestate
transducers, and language modeling.
Selected advanced topics will be covered in
more depth.
COMS E6900x and y Tutorial in computer
science
1–3 pts.
Prerequisite: Instructor’s permission. A reading
course in an advanced topic for a small number
of students, under faculty supervision.
COMS E6901x Projects in computer science
1–12 pts.
Prerequisite: Instructor’s permission. Software
or hardware projects in computer science.
Before registering, the student must submit a
written proposal to the instructor for review.
The proposal should give a brief outline of the
project, estimated schedule of completion, and
computer resources needed. Oral and written
reports are required. May be taken over more
than one semester, in which case the grade
will be deferred until all 12 points have been
completed. No more than 12 points of COMS
E6901 may be taken. Consult the department for
section assignment.
engineering 2011–2012