Development and Evaluation of a Keyword-Accessible Lecture Video Player and Lecture Video Contents

Development and Evaluation of a Keyword-Accessible Lecture Video Player and Lecture Video Contents

Takahiro Yoshida (Tokyo University of Science, Japan) and Seiichiro Hangai (Tokyo University of Science, Japan)
Copyright: © 2009 |Pages: 13
DOI: 10.4018/978-1-59904-480-4.ch008

Abstract

It is desirable to build up a lecture video library to enable students to view past lectures at any time and from anywhere on their PCs. For this purpose, we developed a lecture video player/maker system (Yoshida, 2002, 2003). In developing this system, we considered the usability for students and operability for teachers. The player includes a keyword access function, which enables the student to jump to scenes where one of the registered keywords was spoken. For this purpose, the lecture video maker realizes automatic index generation after continuous speech recognition of the whole lecture stream. In this paper, we discuss the structure and the functions of an ideal lecture video player, and the importance of the index corresponding to the scenes in which the related keywords are spoken. We will also present experimental results regarding keyword extraction from three lecture streams. Evaluations of the lecture videos and the player by students are discussed, and the desirable style of lecture videos for students is surveyed.
Chapter Preview
Top

Introduction

With recent improvements in networking technologies and highly efficient video coding schemes, VOD (video on demand) systems are increasingly being applied in the field of education. A number of such systems, including Classroom 2000 (Abowd, 1999, 2000; Brotherton, 1998) and Intelligent Classroom (Franklin, 1999), have been reported previously. The Classroom 2000 focuses on the multimedia lecture style based on PowerPoint presentations. The Classroom 2000 uses special equipment, such as a digitized chalkboard, video projector, and fixed video camera. Students learn by accessing slides and speech mainly on a Web browser. In contrast, Intelligent Classroom consists of computer-aided technology for recording lecture videos. In this system, the teacher can control the changing of slides simply by giving a verbal command. In recording the lecture, the camerawork is controlled automatically based on the teacher's actions.

In our department, an electronic library of lecture videos has been developed to allow undergraduate students to study not only in the university but also at home. To encourage students to learn on their PCs, a good human interface, such as ease of operation and the facility to jump quickly to any particular scene, is required. On the other hand, for teachers, it is preferable to be able to make such video content without special equipment or requirements. For these purposes, we attempted to address the following requirements in the development of an ideal VOD system for use in education.

  • Adaptive to any lecture style: No special lecture style or equipment should be needed. Only a digital video (DV) camera and operator should be required. The system should be useable with even a traditional lecture style using a blackboard and chalk.

  • Ease of installation and operation: It should be easy to make the digital video stream with an index of keywords. The teacher should connect an IEEE1394 line between a DV camera and a PC, and the keywords and their time base should be automatically extracted and recorded by a speech recognition engine.

  • Quality of lecture videos: The word “quality” has two meanings: that is, picture quality itself and quality of the content. The quality of a lecture video is determined by the clarity of characters and figures on the display and by speech quality from the speakers. The video format and resolution on the screen are standardized. Therefore, skill of the camerawork is very important. The teacher's speech should be recorded on the DV tape synchronously with low noise, because the quality of speech affects both the ease of listening and the performance of speech recognition.

  • Usability for studying: A keyword-accessible lecture video player was developed to assist students in studying under their own initiative.

The following chapters present an overview of the lecture video files and the system developed in this study. The results of evaluation by students are discussed and the desirable style of lecture video contents for students is surveyed.

Top

Recording Lecture Video

In our department, we present “Electrical Circuit” and “Digital Signal Processing” classes, which consist of 12 lectures, to undergraduate students. The details of creating the video files are as follows:

Complete Chapter List

Search this Book:
Reset