Michael N. Wallick

A framework for virtual videography

There are a significant number of events that happen on a regular basis that would be worth preserving on video but for which it is impractical to use traditional video production methods. In this paper, we describe one possible way to inexpensively and unobtrusively capture and produce video in a classroom lecture environment. We discuss the importance of cinematic principles in the lecture video domain and describe guidelines that should be followed when capturing a lecture. We continue by surveying the tools provided by computer vision and computer graphics that allow us to determine syntactic information about images. Finally, we describe a way to combine these tools to create a framework for a Virtual Videography system, one that can automatically generate production quality video. This framework is based on the creation of region objects, a semantically related region of video, despite the fact that we can reliably only gather syntactic information.

Hybrid speaker tracking in an automated lecture room

We present a hybrid speaker tracking scheme based on a single pan/tilt/zoom (PTZ) camera in an automated lecture capturing system. Given that the cameras video resolution is higher than the required output resolution, we frame the output video as a sub-region of the cameras input video. This allows us to track the speaker both digitally and mechanically. Digital tracking has the advantage of being smooth, and mechanical tracking can cover a wide area. The hybrid tracking achieves the benefits of both worlds. In addition to hybrid tracking, we present an intelligent pan/zoom selection scheme to improve the aestheticity of the lecture scene.

A portable solution for automatic lecture room camera management

Rapid advances in technology and decreasing costs have made it possible to attach high resolution video cameras to just about any computer and record the interactions in a lecture room. Additionally, lecture rooms may be outfitted with several cameras for this purpose. However, recording the interactions alone does not create effective video. In this paper we present a method for not only recording, but also editing in real-time, lectures. Unlike previous work, the system is highly portable, allowing quick set-up in various types of lecture rooms. This portability is achieved by using the abstraction of virtual cameramen and physical cameras, and a scriptable interface to the editing rules

Virtual videography

Well-produced videos provide a convenient and effective way to archive lectures. In this demonstration, we present a new way to create lecture videos that possess many of the advantages of well-composed recordings without the cost and intrusion of a video production crew. The videos are produced by an automated system called Virtual Videography that employs the art of videography to mimic videographer-produced videos, while being unobtrusive when recording the lectures. The system uses the data recorded by unattended video cameras and microphones to produce a new edited video as an offline post-process. By producing videos offline, our system can use future information when planning shot sequences and synthesizing new shots. Using syntactic cues gathered from the original video and a novel shot planning algorithm, the system makes cinematic decisions without any semantic understanding of the lecture.

Magic boards

Using a marker or chalkboard allows spontaneity during a presentation or discussion. However, this spontaneity comes at the cost of incorporating visual effects, such as those associated with a computer generated presentation, into the board based presentation. Our goal is to combine the spontaneity of board based presentations with the effects and benefits of a computer generated presentation. In this paper, we present Magic Boards, an approach for augmenting a video containing a presentation with marker or chalkboard. The system finds regions of related writing, through time, on the board. Each region is presented to the author as a single image. The author can then repaint the image with more detail or replace it with a new image or a video/animation. The system replaces the writing on the board with the images and videos provided by the author. The result is a new video in which the board has been augmented with material provided by the author, making it easier to read and understand.

A system for placing videotaped and digital lectures on-line

The idea of distance education, or learning at home has become very popular in recent years. Most major Universities offer classes on-line, and several smaller institutes exclusively offer distance classes. The distance courses can be offered either by mail (videotapes and books) or over the Internet (Web pages). Web-based courses are limited by bandwidth restrictions. This makes complex, interactive presentation of information virtually impossible. Videotapes are able to present complex information, however the production and distribution cost of videotapes makes the idea very prohibitive. This paper presents a system for compressing video lectures for Internet transmission, without losing classroom information. In addition, we present algorithms that improve the interactivity of the presentation by use of optical character recognition and speech recognition technologies.