Syllabus

Course Meeting Times

Lectures: 1 session / week, 1.5 hours / session

Course Text

Schmandt, C. Voice Communication with Computers: Conversational Systems. New York, NY: Van Nostrand Reinhold, 1993. ISBN: 9780442239350.

Note: This book is now out of print, and is provided for downloading on the readings page.

Course Structure

The syllabus of this class will be approximately the table of contents of the book. We will add more recent case studies from the literature and have more to say about topics in spatialized audio, speech skimming, non-speech audio, and computer mediated voice communication.

Rather than a series of formal lectures, this course will be based on your reading for the week and asking questions / discussing in class. Thus it is difficult to predict an exact week-by-week syllabus in advance. This document will evolve during the term to reflect our actual rate of progress. During the last third of the semester, we will optionally cover additional topics in non-speech audio and user interfaces for mobile devices.

Grades and Policies

There will be approximately 8 to 10 problem sets, which will be graded on a scale of 1 (awful) to 5 (great). There will be no exams or projects. Grades will be based on problem set performance, with a half-grade factor for class participation, including doing the readings.

Problem sets are due in class on the due date, and will be accepted only in hard copy (unless prior arrangements have been made). Problem sets will be accepted up until the day they are discussed in class, but submissions after the due date will incur a 1/2 point penalty. (Penalty waived if the instructor has more than two ungraded problem sets not yet returned). You can use whatever word processor or text formatter you want, but except in special cases hand written problem sets are not acceptable. Figures (flow charts, box drawings, etc.) may be hand drawn and labeled.

Calendar

LEC # TOPICS KEY DATES
1 Introduction and genres of conversation Problem set 1 out
2 Components of conversation and speech production

Problem set 1 due

Problem set 2 out

3 Speech production and hearing

Problem set 2 due two days after Lec #3

Problem set 3 out

4 Speech coding

Problem set 3 due

Problem set 4 out

5 Accessing recorded speech Problem set 4 due
6 Speech synthesis Problem set 5 out
7 Review of problem sets and speech recognition intro

Problem set 5 due

Problem set 6 out

8 Hidden Markov models Problem set 6 due
9 Applications of recognition: retrieval in voice docs Problem set 7 out
10 Discourse Problem set 7 due
11 Identity, community, and participation  
12 Presence and being there  
13 Managing interruption