AI EDUCATION IN SCS
CSD Courses
15-229: Multimedia Information Processing, Undergraduate and Masters Level
Raj Reddy, Roger Dannenberg, and Bob Thibadeau, Computer Science
Fall Semester '98; Spring Semester '99
Until recently, most computing tasks dealt with numerical, text, and
symbolic data, and Computer Science has emphasized these discrete data
types. Now, digital representations of audio, video, and images are
common. These new data types are often called "continuous media"
because they represent quantities that vary continuously over time
and/or space. Computers are rapidly becoming the technology of choice
for continuous media production, manipulation, and distribution.
Consequently, an understanding of continuous media is essential for many
modern computing tasks.
Multimedia Information Processing (MMIP) teaches students to work with
continuous media on computers. Students will learn to capture, process,
compress, search, index, store, and retrieve various kinds of continuous
media. The course is team and project oriented. Projects will require
work with audio, scanned images, digital video, and other media, all in
digital form. Readings will provide a conceptual and technical
framework for project work.
The goal of MMIP is to make students comfortable manipulating continuous
media. Students will learn the underlying concepts and be able to apply
their understanding to practical problems such as selecting sample rates
and image resolution, selecting appropriate compression schemes,
creating continuous media for the Web, and using various software tools
to manipulate audio, images, video, and other media.
Prerequisites: 15-212 or equivalent is required. Knowledge of PC and
Windows OS is desirable. Knowledge of Signal Processing is useful, as
is some level of Mathematics beyond Calculus, but neither is a
requirement.
15-381: Undergraduate AI; Graduate or Undergraduate
Manuela Veloso, Fall '98; Andrew Moore, Spring '99, Computer Science
/afs/andrew.cmu.edu/course/15/381/www/home.html
This class is about the theory and practice of Artificial Intelligence.
We will study modern techniques for computers to represent knowledge and
make intelligent (sometimes optimal) decisions. The introduced methods
are applicable throughout a large range of industrial, civil, medical,
financial, robotic and information systems. We will ask questions about
AI systems, such as: How to represent knowledge? How to effectively
generate appropriate sequences of actions: How to deal with
uncertainity in the world? How to learn from experience? How to
process and provide reward and punishment? We expect that by the end of
the course students will have a thorough understanding of the
algorithmic foundations of AI, how probability and AI are closely
interrelated, and a strong appreciation of the big-picture aspects of
developing autonomous intelligent systems. Near the end of the course
we will spend several lectures learning about and discussing some
important current application areas of AI and the bouyant industry of AI
spinoffs.
Incoming students must have already taken an algoirthms course such as
15-211 (and ideally, 15-212).
Foundations and principles of robotic manipulation. Topics include
computational models of objects and motion, the mechanics of robotic
manipulators, the structure of mainipulator control systems, planning
and programming of robot actions.
An introduction to the theory and practice of computer vision, i.e. the
analysis of the patterns in visual images of the world with the goal of
reconstructing the objects and processes in the world that are producing
them. This includes the "low-leve" algorithms of image processing,
multi-scale analysis, segmentation of images, correspondence of multiple
images and reconstruction of depth. It continues with "high-level"
algorithms of pattern recognition and the analysis and recognition of
shapes, objects and scenes using feature, templates and models. The
discussion will be guided by comparison with human and animal vision,
from psychological and biological perspectives. The course should be
appropriate for graduate students in all disciplines and for advanced
undergraduates.
Prerequisites: 15-212; 21-241.
15-780: Graduate Core AI; 16-781: Graduate Core AI (for Robotics students)
Andrew Moore, Computer Science/Robotics
Spring Semester '99
http://www.cs.cmu.edu/~awm/www731
Description:
An introduction to neural networks for computer scientists and
engineers. No previous exposure to the field is assumed. Includes
hands-on experience with a variety of neural net architectures modeled
in MATLAB, and an in-depth look at problems in pattern recognition and
machine learning.
An in-depth study of information processing in real neural systems, from
a computer science perspective. Examines a variety of biological
structures where processing is now sufficiently well understood that it
can be discussed in terms of specific representations and algorithms.
Focuses primarily on computer models of these systems, after
establishing the necessary anatomical, physiological, and psychophysical
context. There are some neuroscience tutorial lectures for those with
no prior background in this area.
15-886/11-741: Information Retrieval (CS core unit / LTI core)
Graduate, Undergraduate (if qualified)
Yiming Yang, Computer Science
Semester: TBA
http://www.cs.cmu.edu/
~yiming/courses/ircore.html
The graduate IR core course focuses on fundamental techniques for
information retrieval, and new research challenges in the field. The
fundamental techniques include: document and query representation,
retrieval models using vector spaces and probabilistic ranking, term
techniques, evaluation measures and methodology. The new challenges
include statistical methods and machining learning techniques applied to
text retrieval, categorization, clustering, summarization, information
extraction and discovery, cross-language and multi-media information
retrieval.
15-889: Continuous Planning, Execution, and Learning; Graduate
Manuela Veloso and Reid Simmons, Computer Science
Spring-99
In this course we will teach the challenges of building autonomous
agents that need to continuously plan, execute their actions, and learn
from their interactions with the environment. In its essence, the
autonomous agents need to "select actions" to achieve its objectives
(and act, and monitor their execution, etc.).
The course will be divided into the following main parts:
I. Deliberative Planning
Static and deterministic domains
Classical planning algorithms
Action representation, abstraction, goals, and heuristics
Learning: domain structure, search efficiency, solution quality
II. Planning under Uncertainty
Dynamic and nondeterministic domains
Conditional planning
Probabilistic planning
Learning: models, iterative planning algorithms, reinforcement learning
III. Plan Execution
Interleaved planning and execution
Reactive planning
Plan monitoring
Replanning
Learning: model refinement, optimal policy, policy reuse
IV. Planning and Execution under Uncertainty
Uncertainty in state, goals, model, and observations
Markov models, POMDPs
Kalman-Bucy filter, Viterbi algorithm, and extensions
Learning: EM algorithm
Evaluation:
This will be a lecture course. There is no textbook, but students will
study research papers. There will be homework, a research project, and
a final exam.
15-xxx and 11-xxx: Language and Information Technologies, Undergraduate
(Junior/Senior) and ~Masters levels
Jaime Carbonell, Computer Science and LTI
Spring Semester '99
This course covers computational approaches for processing natural
language, including: (1) parsing Algorithms, (2) Machine Translation,
(3) Information Retrieval, (4) Learning from text, (5) Coping with
sponateous speech-recognized language, and (6) Applications of these
methods for the Web, newswires, and Digital Libraries. The course will
include aspects of Machine Learning and empirical studies. Hands-on
programming and group projects will be stressed.
15-212 or equivalent is required. Knowledge of AI is useful, as at some
level of Mathematics beyond calculus, but neither is a requirement.
LTI Courses
11-511/11-711: Algorithms for Natural Language Processing; Graduate,
Junior/Senior may participate with instructor approval
Alon Lavie, Bob Frederking, Eric Nyberg, LTI
Fall Semester '98
http://www.cs.cmu.edu/~alavie
Algorithms for NLP is an introductory graduate-level course on the
computational properties of natural languages and the fundamental
algorithms for processing natural languages. The course will provide an
in-depth presentation of the major algorithms used in NLP, including
Lexical, Morphological, Syntactic and Semantic analysis, with the
primary focus on parsing algorithms and their analysis.
11-731/11-531: Machine Translation, Graduate and Undergraduate
Teruko Mitamura, Language Technologies Institute
Semester: TBA
http://www.lti.cs.cmu
.edu/courses/11-731-desc.html
Machine translation is an introductory graduate-level course surveying
history, techniques, and research topics in the field. It is designed
to be taken in conjunction with 11-732, Self-Paced Laboratory in MT.
The main objectives of the course are:
-Obtain a basic understanding of MT systems and MT-related issues.
-Learn about theory and approaches in Machine Translation.
-Learn about basic techniques for MT development, in preparation for the
MT Lab course and real-world MT system project development.
-Obtain in-depth knowledge of one current topic in MT, or
Perform an analysis of a given MT problem, matching it with the most
suitable techniques (includes research, report and presentation).
Prerequisites: 11-721 Grammars and Lexicons (or equivalent background
is recommended).
11-711 Algorithms for NLP (or equivalent background is recommended).
11-751: (A) Speech Recognition and Understanding
Waibel, Reddy, Rudnicky, and others, Computer Science
Semester: TBA
1 CS Coreunit; 12 University Units
The technology to allow humans to communicate by speech with machines or
by which machines can understand when humans communicate with each other
is rapidly maturing. This course provides an introduction to the
theoretical tools as well as the experimental practice that has made the
field what it is today. We will cover theoretical foundations,
essential algorithms, major approaches, experimental strategies and
current state-of-the-art systems and will introduce the participants to
ongoing work in representation, algorithms and interface design.
This course is suitable for graduate students with some background in
computer science and electrical engineering, as well as for advanced
undergraduates.
Prerequisites: This course is primarily for graduate students in LTI,
CS, HCII, Robotics, ECE, Psychology, or Computational Linguistics.
Others by prior permission of instructor. Students should have a
sound mathematical background, knowledge of basic statistics, and good
computing skills. No prior experience with speech recognition is
necessary.
Robotics Courses
16-362 (Graduate) and 16-862 (Undergraduate): Mobile Robot Programming
16-761: Introduction to Mobile Robotics, Graduate and Undergraduate
Illah Nourbakhsh, Robotics
16-362 and 16-862: Fall Semester
16-761: Spring Semester
http.//www.cs.cmu.edu/~illah/
teaching.html
16-362/16-862 This course covers all aspects of mobile robotics
programming, starting at low-level PID control and behavioral control
and graduating all the way to robot team communication and interleaving
planning and execution.
The class presents a strong formal approach and will apply those
formalisms to real robots that you program in teams. We will use six
Nomad Scout robots. This course is for any undergraduate or graduate
who has working knowledge of at least one programming language and has
strong intellectual enthusiasm. This class, which is limited
enrollment, will challenge you.
16-761 Introduction to Mobile Robots is a qualifier for the Robotics
Ph.D. program and is taught in the Spring. This course itroduces Mobile
Roboticvs with historical context, covering traditional and advanced
sensors, effectors and robot control systems.
Automatic planning of manipulator programs based on classical methods of
representing force, motion, and contact.
16-743: Robot Control, Graduate/Undergraduate (admitted w/permission of
instructor)
Chris Paredis, Robotics
Semester: TBA
http://www.cs.cmu.edu/~paredis/16-
743
Introduction of fundamental methodologies for designing feedforward and
feedback controllers for robot manipulators. The course covers a wide
variety of control disciplines (kinematic control, optimal control,
robust control, adaptive control, etc.) and focuses on how these
approaches can be applied to the specific case of robot manipulator
control. It introduces the students to general methodologies for
analyzing stability and performance of control algorithms, and provides
them with practical experience designing linear and non-linear
controllers.
The lure of better performance, lower cost, and miniaturization of
sensor and actuator systems has motivated growth in the area of
silicon-based integrated microsystems. Microsystems technology has
broad applications such as miniature inertial measurement units,
biochemical analysis on a chip, arrayed micromanipulation of parts,
optical displays, and micro-probes for neural recording. This course is
an introduction to microsensor and microactuator technology, intended
for first and second-year graduate students in ECE and Robotics. The
course provides the engineering background necessary for design of
integrated microsystems. Homework and exams reinforce the engineering
material taught in class. Students will research a chosen topic and
design a microsystem as part of the final project. The project will
take the form of a journal paper accompanied by an oral presentation to
the class.
Course topics include: substrate (bulk) micromachining; surface
micromachining; high-aspect-ratio processes; thin-film properties;
micromechanical design; micromechanical fabrication services;
electromechanical force; electrostatic, piezoelectric, and thermal
actuation; capacitive, magnetic, thermal, chemical, and biological
sensing mechanisms; sensing circuits; noise sources, pressure sensors;
inertial sensors; tactile imagers; resonant sensors; microfluidics;
microrobotics; micro-optics; computer-aided design, modeling, and
simulation.
16-879/18-879/24-700A: Mechatronic Design, Graduate and Undergraduate
Gary Fedder, Howie Choset (cross listed as ECE/MechE/RI course)
Spring Semester '99
http://www.ece.cmu.edu/afs/e
ce/class/ee879
Mechatronics is the synergistic integration of mechanism, electronics,
and computer control to achieve a functional system. Because of the
emphasis upon integration, this course will center around laboratory
projects in which small teams of students will configure, design, and
implement a succession of mechatronic subsystems, leading to system
integration in a final project. Lectures will complement the laboratory
experience with comparative surveys, operational principles, and
integrated design issues associated with the spectrum of mechanism,
electronics, and control components. The course is open to ECE, ME, and
RI graduate students and advanced undergraduates. Class size is limited
to 30 students, split evenly among the three departments.