-
Notifications
You must be signed in to change notification settings - Fork 0
/
cs294-video.html
202 lines (158 loc) · 9.98 KB
/
cs294-video.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
<head>
<TITLE>Video Conferencing</TITLE>
</head>
<body bgcolor = "FFFFFF">
<h1>Video Conferencing</h1>
<h3>discussion led by <a href="http://www.cs.berkeley.edu/~abegel">Andrew Begel</a>
and <a href="http://www.eecs.berkeley.edu/~michaels">Michael Shilman</a></h3>
<hr>
<h2>Papers Read</h2>
<ul>
<li> Egido, Carmen. <i>Video Conferencing as a Technology to Support Group Work: A Review of its
Failures.</i> In CSCW '88.
<li> Macedonia, Brutzman, <i>"MBone Provides Audio and Video Across the Internet",</i> IEEE
Computer, Apr 94, v27n4, p.30-36
<li> Brittan, David. <i>Being There: The Promise of Multimedia Communications.</i> In MIT
Technology Review, pp. 42-50. May/June 1992.
<li> Curtis, Pavel, Dixon, Michael, Frederick, Ron, and Nichols, David A. <i>The Jupiter Audio/Video
Archtecture: Secure Multimedia in Network Places.</i>
In Multimedia '95, pp. 79-90. </ul>
<hr>
<h3>Video Conferencing as a Technology to Support Group Work:
A Review of its Failures.</h3>
This paper was written in 1988, before any of the current MBone
tools and video conferencing sessions became widely available and accessible.
The authors, from Bell Communications Research, describe the history of the
video conferencing movement from the first idea of it back in the 1930's all
the way up to the spectacular failure of the videophone in the late 1970's.<p>
They attribute this failure to a culture problem. There is a fear of embarrassment
from appearing in front of the video phone in your pajamas, hair all unkempt,
looking like a slob. Using an audio phone, people can mask any aspect of their
physical appearance, since it takes far less work to make your voice sound good
than to spend time on your hair. This phone seemed like an invasion of privacy.<p>
We thought another problem was due to the fact that most Americans are conservative
and luddites at heart. New technology scares people to the point where they
won't try again at something if it failed before. It has to be presented in a
new way. AT&T's videophone was simply a better version of the videophone that everyone
had heard so much about (and never seen) in the 60's. They remembered all the bad
things about the videophone and never even considered any of its good qualities that AT&T
tried to stress (Showing your grandkids to their grandparents by video was one popular
image). <p>
In addition to perceptual problems, there was a problem of distribution. How many
people do <b>you</b> know who have a videophone? Yes, exactly the problem. Without
a <i>critical mass</i> of videophone purchasers, consumers basically got a really
expensive phone whose video display remained blank most of the time. <p>
Michael Shilman thought about the difference between the American culture and the
Japanese culture. The Japanese tend to be gadgetophiles. They buy any new electronic
toy that comes on the market. Take Sony minidisc. The technology is smaller, lightweight,
recordable and better than tape. Why do most Japanese have really small Sony minidiscmans,
while American's are still clunking around with huge fullsize discmans? Americans didn't
see enough of an improvement over the fullsize disc to justify buying it. Most people
remember the big debate when 8-track gave way to the cassette tape. They also remember
LPs giving way to tapes and CDs. They had no intention of throwing out their amassed
assets just to buy them over again with no appreciable difference in quality other than
the minidiscs not holding as much as the fullsize discs.<p>
Another aspect of this paper addresses why people want video conferencing in the first
place. It claims results from several studies (with questionable methodologies)
over a number of years that show that the debate is almost over. In the beginning,
people were excited about the technology and predicted that 85% of future meetings
would be done over video. Further studies showed that this number dropped over
time until about 4% today.<p>
Michael Shilman also came up with a graph of communication styles from IRC talk to
face-to-face conversation that can be found on the
<a href="http://www.cs.berkeley.edu/~landay/cscw/lectures/video/video-mb.ps">
Media Board archives.</a><p>
<hr>
<h3>MBone Provides Audio and Video Across the Internet</h3>
This paper gave a simple description of what the MBone is. The people who wrote the paper
were pretty enthusiastic about this great new technology (the paper was from 1994, so I don't
know where they've been for the past few years). It discusses the network concepts behind the
MBone, bandwidth considerations, MBone tools, events and interesting MBone uses. <p>
Perhaps the most interesting line in this paper was, <b>"It was just like being there."</b>
We found this amusing at the time, since we were presenting these papers live over the MBone
to the class. Staring into a video camera, instead of looking at the monitor where the
pictures of the people were, was not "just like being there." In addition, since we were the
focus of the conversation, conversation tended to die unless we kept it going. There was none
of the subtle eye contact or body language that lets other people know that you want them to talk
now. <p>
<hr>
<h3>Being There: The Promise of Multimedia Communications</h3>
<i>Being There: The Promise of Multimedia Communication</i> gives an overview
of several videoconferencing systems in development at research labs.
For each system, it describes the mode of communication, the way in
which communication is initiated, and includes several comments from the
developers. The systems described include AT&T's Rapport, NTT's Team
WorkStation, Xerox PARC's Media Space, Video Draw, and Video Whiteboard,
Bellcore's Cruiser, and Olivetti's Pandora.<p>
In Rapport, the key metaphor is a "virtual meeting room" in which a user
enters the room and call other people to join in. It provides a mixture
of A/V and document communication, in which the A/V takes a back seat to
other modes of discourse (specifically the document sharing). One
person can use the document editing software at a time, and users
"raises" a virtual hand (cursor) to gain editing privileges. The system
also has a "store and forward" function which allow meetings to be saved
and passed on to other people.<p>
Team WorkStation is an "open shared workspace" which uses video as its
primary medium. User's sketches are captured on video and are
superimposed on one another to achieve this effect. Ishii's current
work involves making the UI disappear completely from the collaborative
experience.<p>
Media Space views video as the most important aspect of the
collaborative design, with the claim that design is more of a social
task than a technical one. The ambiguity of being able to talk on video
allows for more expressive freedom in the design process. Work here
lead to VideoDraw, which combines a sketching surface with a video
screen. PARC researchers note that users prefer more manual tasks in
the collaborative environments, and hypothesize that this is a signal
their system is not quite right yet.<p>
Cruiser is a "casual encounter" based video system. Users can make
quick calls into offices to see if people are around. Users typically
make only short conversations on the system, and often pack up their
stuff and meet in person for anything longer than a couple minutes.
Bellcore has also developed VideoWindow, which is a large virtual window
separating parties at different sites. They note that it is often easy
to ignore people on the other side of the window.<p>
<hr>
<h3>The Jupiter Audio/Video Archtecture: Secure Multimedia in Network Places</h3>
Jupiter is a multi-user, multimedia virtual world built on top of MOO
(Object-Oriented MUD). It incorporated audio and video objects that
people could program in various ways, subject to system
constraints. There were quite a few important ideas in this paper
about privacy, security, flexibility, archiving, playback and
awareness.<p>
There were also some interesting MOO applications and suggestions for
future applications:<p>
<ul>
<li> Synchronous communication between people in the same room.
<li> Psst. Private audio/visual communication.
<li> Ambient Noise -- Room reactions. (background noise, physical->audio effects)
<li> Noise from other rooms. Public audio leakage.
<li> Video conferencing.
<li> Workgroup media spaces. Chat rooms with context.
<li> P/A system.
<li> Videophone/answering machine. Answering machine could be an intelligent agent.
<li> IRC Phone
<li> VCR and tape recorder (how will they do indexing?)
</ul>
The people who wrote this system brought their MOO privacy/security paranoia to
Jupiter. They include end-to-end encryption of all A/V streams, and switch keys
everytime someone enters or leaves a room. All objects have inherent permissions on
ownership, modification, examination, etc. <p>
Like all good secure systems, the fail-safe mode is denial-of-access. There is no
unintentional video send or receive with the users explicit go ahead. There will be
no Big Brother watching over you unless you allow it. <p>
How can this system deal video archival and playback? Each segment may have a
different encryption and security level, and thus may allow people to see only
small portions of it. If you have video playback, how can you tell it's not
live? Could it be a bot playing back prerecorded video snippets? Is it live or
is it Memorex? Does it matter? If you expose the source of the video, is it
in violation of the source's privacy or anonymity? What if the source is a bot?
Is it violation of the bot's privacy? <p>
<Hr>
<address>Andrew Begel
<a href="mailto:[email protected]">[email protected]</a></address><br>
<address>Michael Shilman
<a href="mailto:[email protected]">[email protected]</a></address><p>
<!-- hhmts start -->
<!-- hhmts end -->
</body>