WEBVTT

1
00:00:27.660 --> 00:00:30.840
For this lecture we will take a look at some motion capture videos

2
00:00:30.840 --> 00:00:33.759
to see how to install a motion capture,

3
00:00:33.759 --> 00:00:35.279
how it is filmed

4
00:00:35.279 --> 00:00:37.480
and the developing

5
00:00:37.480 --> 00:00:40.400
technology flows

6
00:00:40.400 --> 00:00:42.360
We will also look at how to

7
00:00:42.360 --> 00:00:46.799
work together with other equipments or other fields

8
00:00:46.799 --> 00:00:49.480
while shooting motion capture

9
00:00:49.480 --> 00:00:51.880
Also we will look into how contents are

10
00:00:51.880 --> 00:00:54.480
being expanded and used these days

11
00:00:54.480 --> 00:00:57.960
Progress of Motion Capture Technology

12
00:00:58.400 --> 00:01:00.799
This video is about the Magnetic system

13
00:01:00.799 --> 00:01:04.040
Nowadays, there are lots of equipment

14
00:01:04.040 --> 00:01:06.080
using gyro sensors

15
00:01:06.080 --> 00:01:09.680
Gyro sensors can measure not only

16
00:01:09.680 --> 00:01:13.199
rotation data, but also speed and acceleration

17
00:01:13.199 --> 00:01:15.160
It's also wallet-friendly

18
00:01:15.160 --> 00:01:17.360
So it is trending as of now

19
00:01:17.360 --> 00:01:19.239
The video you are about to see is

20
00:01:19.239 --> 00:01:22.480
a magnetic system using gyro sensors

21
00:01:22.480 --> 00:01:26.279
You can see it utilizes a large space

22
00:01:26.279 --> 00:01:28.519
I will explain after the video

23
00:01:29.745 --> 00:02:08.139
Magnetic System
- Xsens MVN

24
00:02:08.880 --> 00:02:10.880
Magnetic sensors in the past was

25
00:02:10.880 --> 00:02:14.320
vulnerable to high voltage currents and rebar material

26
00:02:14.320 --> 00:02:16.080
Filming done in

27
00:02:16.080 --> 00:02:18.880
wooden buildings were not a

28
00:02:18.880 --> 00:02:20.880
big problem

29
00:02:20.880 --> 00:02:23.919
However it was very difficult

30
00:02:23.919 --> 00:02:26.380
to film near concrete buildings with rebar

31
00:02:26.400 --> 00:02:30.039
to use the magnetic system

32
00:02:30.039 --> 00:02:34.160
Many of these issues have been

33
00:02:34.160 --> 00:02:37.020
resolved by using the gyro sensor

34
00:02:37.020 --> 00:02:42.199
but we still have to deal with some data errors

35
00:02:42.199 --> 00:02:45.220
regarding current flow and reinforcing bar material

36
00:02:45.220 --> 00:02:49.839
However, as software resolves these issues

37
00:02:49.839 --> 00:02:53.100
the frequency of use is increasing

38
00:02:53.100 --> 00:02:54.880
In the video

39
00:02:54.880 --> 00:02:59.039
you saw that we were able to use a large space freely

40
00:02:59.039 --> 00:03:03.299
And the data comes out seemlessly

41
00:03:03.299 --> 00:03:06.080
even though there are steel material all around the set

42
00:03:06.814 --> 00:03:39.444
Magnetic System
- Xsens MVN

43
00:03:40.740 --> 00:03:44.000
Here you can see that the data for

44
00:03:44.000 --> 00:03:46.240
driving a car and riding a bike

45
00:03:46.240 --> 00:03:49.580
have been extracted well

46
00:03:49.580 --> 00:03:52.099
Like this, the magnetic system's data

47
00:03:52.099 --> 00:03:55.140
have been greatly improved by the use of gyro sensors

48
00:03:55.140 --> 00:03:57.819
and we can use it with less difficulties

49
00:03:57.819 --> 00:04:00.039
There are still some areas where

50
00:04:00.039 --> 00:04:02.139
the legs are blurred and the hands are out of sync

51
00:04:02.139 --> 00:04:05.319
while shooting in real time

52
00:04:05.319 --> 00:04:09.000
but it is still widely used in SNS content using

53
00:04:09.000 --> 00:04:12.240
virtual humans and games and animation production

54
00:04:12.240 --> 00:04:15.939
that can extract a large amount of data in smaller amounts of time

55
00:04:15.939 --> 00:04:18.620
Games, for example

56
00:04:18.620 --> 00:04:23.279
We apply a technique called Motion Matching

57
00:04:23.279 --> 00:04:25.440
This technique

58
00:04:25.440 --> 00:04:30.000
make the object run 20 to 30 meters

59
00:04:30.000 --> 00:04:32.880
and extracts the data

60
00:04:32.880 --> 00:04:36.480
for the required section

61
00:04:36.480 --> 00:04:40.880
When these techniques use Optical systems

62
00:04:40.880 --> 00:04:44.000
We need area for installation

63
00:04:44.000 --> 00:04:46.000
and it will take a lot of time

64
00:04:46.000 --> 00:04:49.200
But as for Gyro systems

65
00:04:49.200 --> 00:04:51.519
If we do not have an indoor area

66
00:04:51.519 --> 00:04:56.640
we can shoot at outdoor areas

67
00:04:56.640 --> 00:04:59.390
So the Gyro systems are

68
00:04:59.390 --> 00:05:01.119
preferred in the game scene

69
00:05:01.119 --> 00:05:04.239
Next we will look in to the Optical system

70
00:05:04.239 --> 00:05:08.720
This video shows the Optical systems

71
00:05:08.720 --> 00:05:11.820
default settings, software installations

72
00:05:11.820 --> 00:05:14.000
network settings and camera installations

73
00:05:14.000 --> 00:05:18.399
Also briefly shows the process of calibration

74
00:05:18.399 --> 00:05:23.040
attaching body and tool markers on the character

75
00:05:23.040 --> 00:05:25.519
and converting them to third-party tools

76
00:05:25.519 --> 00:05:30.320
Ultimately, you can see how it is implemented

77
00:05:30.320 --> 00:05:32.799
in Motion Builder, Unity, Unreal Engine, etc.

78
00:05:32.799 --> 00:05:35.359
Let's go ahead to the video

79
00:05:35.964 --> 00:06:13.128
Optical System
- Optitrack

80
00:06:14.007 --> 00:06:19.208
Optical systems require the hassle of installing cameras

81
00:06:19.208 --> 00:06:22.575
because data must be created by the camera

82
00:06:22.575 --> 00:06:25.575
But once installed and calibrated

83
00:06:25.575 --> 00:06:29.255
you can produce optimal data, as long as you do not withdraw

84
00:06:29.255 --> 00:06:31.655
For most studios you can

85
00:06:31.655 --> 00:06:34.284
continue to use after the first installation

86
00:06:34.284 --> 00:06:38.657
But in the instance of overseas or large space shooting

87
00:06:38.677 --> 00:06:40.677
you would usually rent a large studio

88
00:06:40.677 --> 00:06:44.257
about 30 to 50 meters

89
00:06:44.271 --> 00:06:46.451
and use it for a month or two

90
00:06:46.451 --> 00:06:49.259
and move out

91
00:06:49.259 --> 00:06:51.760
Because it would be very costly

92
00:06:51.760 --> 00:06:54.239
to keep the space and equipment for longer

93
00:06:54.239 --> 00:06:57.839
These types of projects are

94
00:06:57.839 --> 00:07:00.480
gradually emerging in Korea as well

95
00:07:00.480 --> 00:07:02.239
The chart here shows

96
00:07:02.239 --> 00:07:06.799
the difference between Gyro and Optical systems

97
00:07:06.799 --> 00:07:10.160
the Gyro system does not have space limitations

98
00:07:10.160 --> 00:07:13.359
and can be used indoors

99
00:07:13.359 --> 00:07:16.880
so it has a big advantage in terms of location

100
00:07:16.880 --> 00:07:19.679
The optical system has space limitations

101
00:07:19.679 --> 00:07:21.679
it should be shot indoors

102
00:07:21.679 --> 00:07:24.239
and also needs time for installations

103
00:07:24.239 --> 00:07:27.839
gyro systems do not need post-edits

104
00:07:27.839 --> 00:07:30.799
It has the advantage of saving

105
00:07:30.799 --> 00:07:33.600
a lot of time and money

106
00:07:33.600 --> 00:07:36.640
Gyro systems have the disadvantage of

107
00:07:36.640 --> 00:07:39.760
lower data quality compared to optical systems

108
00:07:39.760 --> 00:07:41.679
due to low accuracy

109
00:07:41.679 --> 00:07:45.279
Optical systems need post-edits

110
00:07:45.279 --> 00:07:48.959
so it can cost you both time and money

111
00:07:48.959 --> 00:07:51.440
but compared to the gyro it has

112
00:07:51.440 --> 00:07:53.440
the upper-hand in data quality

113
00:07:53.440 --> 00:07:56.320
with the 0.1mm accuracy

114
00:07:56.320 --> 00:07:59.359
Therefore it is suitable for works requiring

115
00:07:59.359 --> 00:08:02.860
quick movements, detailed acting and accurate measurements

116
00:08:03.600 --> 00:08:08.079
The video about to be played uses motion capture

117
00:08:08.079 --> 00:08:11.279
it is a 2014 film called Big Match

118
00:08:11.279 --> 00:08:13.839
It was used as a previs

119
00:08:13.839 --> 00:08:18.619
This was probably the first time in Korean cinema

120
00:08:18.619 --> 00:08:20.799
that motion capture has been used from previs

121
00:08:20.799 --> 00:08:23.040
Let's watch the video first

122
00:08:23.618 --> 00:08:32.919
Previs and Motion Capture used in 'Big Match'

123
00:08:32.919 --> 00:08:36.559
I took the role in this film for

124
00:08:36.559 --> 00:08:40.400
the martial arts, special effects, CG, etc.

125
00:08:40.400 --> 00:08:44.880
to be more intensely and elaborately

126
00:08:44.880 --> 00:08:46.880
designed into the film

127
00:08:46.880 --> 00:08:50.719
I attempted to add precision to the movements by

128
00:08:50.719 --> 00:08:55.919
motion-capturing the action sequences in advance

129
00:08:55.919 --> 00:08:59.599
those techniques were used for things like Ikho's digital character

130
00:08:59.599 --> 00:09:01.099
At the police station scene

131
00:09:01.099 --> 00:09:04.400
Ik-ho was simply curious about his brother

132
00:09:04.400 --> 00:09:06.719
and he had to do as Ace told him to do so

133
00:09:06.719 --> 00:09:09.359
and get out without harming the police

134
00:09:09.359 --> 00:09:11.609
I think this can be

135
00:09:11.609 --> 00:09:12.694
the mission and fun of this movie

136
00:09:12.714 --> 00:09:14.894
Perfected by 3D
Mission Game in the Middle of the City

137
00:09:14.894 --> 00:09:19.244
The setting with most excitement and CG effert

138
00:09:19.244 --> 00:09:20.799
is of course the Sangam Stadium

139
00:09:20.799 --> 00:09:23.840
Because we were trying to use the entire stadium

140
00:09:24.380 --> 00:09:28.960
previs and motion capture technology was incorporated

141
00:09:28.960 --> 00:09:34.320
to design each sequence more accurately

142
00:09:34.320 --> 00:09:38.239
including the precision of movements, actors' routes

143
00:09:38.239 --> 00:09:39.939
and camera angles for the main character

144
00:09:39.939 --> 00:09:42.640
Ik-ho's action scenes

145
00:09:42.640 --> 00:09:44.340
At the time, the concept of

146
00:09:44.340 --> 00:09:47.359
previs was not well known in Korea

147
00:09:47.359 --> 00:09:51.039
So this might be the very first Korean film to

148
00:09:51.039 --> 00:09:54.239
try Previs using motion capture

149
00:09:54.239 --> 00:09:58.460
In other countries, previs technology

150
00:09:58.460 --> 00:10:01.560
has been used for over 20 years

151
00:10:01.560 --> 00:10:03.610
Previs is gradually earning its

152
00:10:03.610 --> 00:10:06.520
popularity recently

153
00:10:06.520 --> 00:10:08.420
Motion capture technology also

154
00:10:08.420 --> 00:10:12.420
is being used more and more

155
00:10:12.420 --> 00:10:15.500
The next video is a Korean video

156
00:10:15.500 --> 00:10:17.700
The children in the first scene

157
00:10:17.700 --> 00:10:21.020
are digital characters made by motion capture

158
00:10:21.020 --> 00:10:24.620
The setting is impossible to find in real life

159
00:10:24.620 --> 00:10:27.320
and even if found, it would be too dangerous

160
00:10:27.320 --> 00:10:29.860
to bring the children there to shoot

161
00:10:29.860 --> 00:10:32.520
Hence it was made digital

162
00:10:32.520 --> 00:10:36.040
This probably made in the early 2000

163
00:10:36.040 --> 00:10:37.860
Motion capture technology would be necessary

164
00:10:37.860 --> 00:10:41.540
in order to create videos where characters

165
00:10:41.540 --> 00:10:44.040
must move in a background that is not real

166
00:10:44.040 --> 00:10:46.340
Next is an edited video of the

167
00:10:46.340 --> 00:10:49.840
digital characters in a Korean drama

168
00:10:49.840 --> 00:10:53.040
Digital characters are being used efficiently

169
00:10:53.040 --> 00:10:55.240
in the large crowd scene

170
00:10:55.240 --> 00:10:57.119
Let's watch the video

171
00:10:57.156 --> 00:12:23.179
Motion Capture Uses in Korea

172
00:12:24.310 --> 00:12:27.940
For the drama, digital characters were created first

173
00:12:27.940 --> 00:12:30.419
then the movements were made with motion capture

174
00:12:30.419 --> 00:12:33.419
and then it was turned into

175
00:12:33.419 --> 00:12:35.619
Database for use

176
00:12:35.619 --> 00:12:40.580
The first scene is an action scene from the drama ‘Gu Family Book’

177
00:12:40.580 --> 00:12:44.179
Enemies would turn in to ash when defeated

178
00:12:44.179 --> 00:12:48.179
so all of them had to be made from digital characters

179
00:12:48.179 --> 00:12:51.539
and motion capture was used for more realistic movements

180
00:12:51.539 --> 00:12:56.259
Also in dramas like ‘The Moon Embracing the Sun’

181
00:12:56.259 --> 00:13:00.100
‘The King's Doctor’, and ‘The God of War’

182
00:13:00.100 --> 00:13:03.300
the smaller characters at the outer angle

183
00:13:03.300 --> 00:13:07.059
are made with digital characters in means of

184
00:13:07.059 --> 00:13:10.179
reducing costs efficiently

185
00:13:10.179 --> 00:13:15.059
Like for the performance scene in 'Light and Shadow'

186
00:13:15.059 --> 00:13:19.300
Supporting cast were placed in the front of the audience

187
00:13:19.300 --> 00:13:21.460
and for the back

188
00:13:21.460 --> 00:13:24.419
digital characters were used

189
00:13:24.419 --> 00:13:29.059
We collected about 10 cheering gestures during the motion capture

190
00:13:29.059 --> 00:13:31.619
ran a crowd simulation

191
00:13:31.619 --> 00:13:34.518
to create an audience over 1,000 people

192
00:13:34.518 --> 00:13:38.118
Collaboration and Application with Other Fields

193
00:13:38.660 --> 00:13:41.779
Let's move on to the next case

194
00:13:41.779 --> 00:13:45.059
This is a 2007 american film called Beowulf

195
00:13:45.059 --> 00:13:47.139
It is almost 17 years old

196
00:13:47.139 --> 00:13:51.539
but had already incorporated today's hi-technology

197
00:13:51.539 --> 00:13:54.259
It used 240 cameras

198
00:13:54.259 --> 00:13:58.339
There were about 200 of markers

199
00:13:58.339 --> 00:14:01.300
per person on the body, face, fingers and etc.

200
00:14:01.300 --> 00:14:03.940
It filmed 10 people simultaneously

201
00:14:03.940 --> 00:14:07.539
It also used various of different structures

202
00:14:07.539 --> 00:14:10.100
and also did motion capture on horses

203
00:14:10.100 --> 00:14:13.139
You can also spot that

204
00:14:13.139 --> 00:14:16.660
virtual camera was used

205
00:14:16.660 --> 00:14:18.705
Let's watch the video

206
00:14:18.755 --> 00:14:47.816
Motion Capture used in 'Beowulf'

207
00:14:48.363 --> 00:14:52.403
This movie has left many records at the time of production

208
00:14:52.403 --> 00:14:56.443
For example, filming was completed in 30 days

209
00:14:56.443 --> 00:15:00.563
It is very unusual to finish shooting

210
00:15:00.563 --> 00:15:02.883
a movie in 30 days

211
00:15:02.883 --> 00:15:05.162
It usually takes 3 to 6 months

212
00:15:05.162 --> 00:15:08.322
but this one only took 30 days

213
00:15:08.322 --> 00:15:11.242
Another one is that the main cast

214
00:15:11.242 --> 00:15:13.963
lived in a trailer next to the set during his shoot

215
00:15:13.964 --> 00:15:17.364
He also limited the scene transitions to

216
00:15:17.364 --> 00:15:18.899
15 minutes

217
00:15:18.899 --> 00:15:22.500
So all scene transitions were done in 15 minutes

218
00:15:22.500 --> 00:15:24.900
Which is probably why the whole shoot

219
00:15:24.900 --> 00:15:26.118
was able to be completed in 30 days

220
00:15:26.118 --> 00:16:08.288
Virtual Camera used in 'Beowulf'

221
00:16:09.059 --> 00:16:13.619
you could see an example of using a virtual camera

222
00:16:13.619 --> 00:16:17.369
In the past, virtual cameras were used

223
00:16:17.369 --> 00:16:19.859
while watching the action in the post studio

224
00:16:19.859 --> 00:16:24.209
But now we can apply motion capture

225
00:16:24.209 --> 00:16:27.700
in real time during the actual filming

226
00:16:27.700 --> 00:16:31.940
I will explain this part later in a different video

227
00:16:31.940 --> 00:16:33.640
The movie 'Beowulf'

228
00:16:33.640 --> 00:16:36.820
used yet another hi-technology

229
00:16:36.820 --> 00:16:40.870
Which is tracking the movement of the pupil using a sensor

230
00:16:40.870 --> 00:16:43.820
This is called a EGO sensor

231
00:16:43.820 --> 00:16:47.940
and it tracks the muscles around the pupil and generates a data

232
00:16:47.940 --> 00:16:50.390
This is a medical sensor

233
00:16:50.390 --> 00:16:53.220
but was applied in movie making

234
00:16:53.220 --> 00:16:55.970
So you can see the slightest shakings

235
00:16:55.970 --> 00:16:57.859
just like real pupils

236
00:16:57.859 --> 00:17:00.079
Let's watch the video

237
00:17:00.115 --> 00:17:36.717
EOG Sensor Used in 'Beowulf'

238
00:17:37.299 --> 00:17:40.820
Next is a 2009 film, 'Avatar'

239
00:17:40.820 --> 00:17:45.619
160 motion capture cameras were used in this movie

240
00:17:45.619 --> 00:17:48.579
The facial captures were filmed

241
00:17:48.579 --> 00:17:52.500
using camera tracking rather than a marker method

242
00:17:52.500 --> 00:17:55.859
It used virtual cameras in real time

243
00:17:55.859 --> 00:17:59.009
and also used a fusion camera that

244
00:17:59.009 --> 00:18:00.660
produces 3D images

245
00:18:00.660 --> 00:18:03.859
'Avatar' not only uses motion capture

246
00:18:03.859 --> 00:18:09.380
but also used Simulcam in real time which is

247
00:18:09.459 --> 00:18:15.140
the combination of virtual and 3D cameras

248
00:18:15.140 --> 00:18:17.025
Let's watch the video

249
00:18:17.025 --> 00:20:15.826
Motion Capture Application Example
'Avatar'

250
00:20:17.380 --> 00:20:20.900
When 'Avatar' was released

251
00:20:20.900 --> 00:20:24.580
there was a huge boom of 3D movies

252
00:20:24.580 --> 00:20:27.940
There was also a lot of recognition of motion captures

253
00:20:27.940 --> 00:20:32.099
Before then it was not too popular in Korea

254
00:20:32.099 --> 00:20:35.059
but the movie 'Avatar' played a big role in its popularity

255
00:20:35.059 --> 00:20:38.260
Interest in Face has increased as well

256
00:20:38.260 --> 00:20:42.659
The marker method had good quality data

257
00:20:42.659 --> 00:20:44.739
but it was very difficult to

258
00:20:44.739 --> 00:20:46.739
apply to a digital face

259
00:20:46.739 --> 00:20:50.900
this was resolved using the camera tracking equipment

260
00:20:50.900 --> 00:20:54.979
In fact, the face camera is

261
00:20:54.979 --> 00:20:58.179
more outdated than the marker method

262
00:20:58.179 --> 00:21:02.500
But its quality has been verified through the 'Avatar'

263
00:21:02.500 --> 00:21:06.419
and it was also easier to apply to digital characters

264
00:21:06.419 --> 00:21:09.140
Hence we use this more and more

265
00:21:09.140 --> 00:21:12.500
and has become established through continuous technological advancement

266
00:21:12.500 --> 00:21:15.380
So now the marker method is gone

267
00:21:15.380 --> 00:21:18.500
as the camera tracking methods develop

268
00:21:18.500 --> 00:21:20.099
now it has evolved into 4D capture method

269
00:21:20.099 --> 00:21:23.859
using multiple cameras

270
00:21:23.859 --> 00:21:27.700
The 4D capture method uses 4 or more cameras

271
00:21:27.700 --> 00:21:31.299
to track the facial expressions

272
00:21:31.299 --> 00:21:33.380
This allows to extract modeling data

273
00:21:33.380 --> 00:21:35.859
of the facial expressions

274
00:21:35.859 --> 00:21:37.140
I will touch on 4D captures

275
00:21:37.140 --> 00:21:40.020
later once again

276
00:21:40.020 --> 00:21:44.099
This video shows the latest

277
00:21:44.099 --> 00:21:45.539
motion capture shooting methods

278
00:21:45.539 --> 00:21:49.539
You will be able to see optical system

279
00:21:49.539 --> 00:21:52.020
and face captures using markers

280
00:21:52.020 --> 00:21:54.340
and making high quality games

281
00:21:54.340 --> 00:21:55.619
using various of structures

282
00:21:55.619 --> 00:21:57.469
Let's watch the video

283
00:21:58.020 --> 00:23:57.070
Performance Capture
'Goodbye Kansas Studio'

284
00:23:57.859 --> 00:24:00.659
In the next video you will see a good example of

285
00:24:00.659 --> 00:24:03.259
virtual production and motion capture usage

286
00:24:03.259 --> 00:24:07.109
This will be a comprehensive collection of technologies that are currently in use

287
00:24:07.109 --> 00:24:09.709
Let's continue after the video

288
00:24:09.762 --> 00:25:27.535
Case of Virtual Production and Motion Capture 
'ABBA Virtual Concert'

289
00:25:28.460 --> 00:25:32.690
This is a 4D capture method that reproduced the youthful appearances

290
00:25:32.690 --> 00:25:35.790
and expressions of members over 70 years old

291
00:25:35.790 --> 00:25:38.690
It shapes up natural gestures and movements

292
00:25:38.690 --> 00:25:41.440
using motion capture technology

293
00:25:41.440 --> 00:25:44.790
And it used LED and holograms

294
00:25:44.790 --> 00:25:48.490
to build a virtual production for this performance

295
00:25:48.490 --> 00:25:52.340
So we moved away from the old method of

296
00:25:52.340 --> 00:25:54.990
production content through post-editing

297
00:25:54.990 --> 00:25:59.340
to a real-time filming with the advancement in technology

298
00:25:59.340 --> 00:26:02.340
In addition, by incorporating various technologies

299
00:26:02.340 --> 00:26:05.940
it is changing into an efficient production method

300
00:26:05.940 --> 00:26:07.540
called virtual production

301
00:26:07.540 --> 00:26:10.090
The key point of virtual productions is

302
00:26:10.090 --> 00:26:12.440
to minimize location shooting

303
00:26:12.440 --> 00:26:15.340
and to do all shooting in real time

304
00:26:15.340 --> 00:26:20.540
making efficient use of time and reducing costs

305
00:26:20.540 --> 00:26:24.590
This concludes the motion capture lecture

306
00:26:24.618 --> 00:26:27.518
Magnetic system
Gyro sensor can measure rotation data, speed and acceleration
Gyro sensors are inexpensive and can be used efficiently

307
00:26:27.568 --> 00:26:29.915
In Korea, most concrete building had rebar, which was difficult to use magnetic System
Recent software solves those problems, resulting in increased usage

308
00:26:29.915 --> 00:26:32.582
Optical System
Can continuously produce data once installed
One-off Optical system projects are trending

309
00:26:32.582 --> 00:26:34.682
Collaboration and Application with Other Fields
'Beowulf': used 240 cameras, 200 markers per person

310
00:26:34.682 --> 00:26:36.782
Able to track eye movements and even small tremors with EOG sensors

311
00:26:36.782 --> 00:26:38.583
'Avatar' Case
160 motion capture cameras used, face capture using marker camera tracking
Virtual cameras used in real time, incorporated fusion camera that produces 3D images

312
00:26:38.633 --> 00:26:40.633
Real-time production using a method called Simul Cam through collaboration between virtual camera and 3D Fusion Camera