1
00:00:00,000 --> 00:00:07,320
[LTA MUSIC]

2
00:00:11,440 --> 00:00:17,280
LiveTextAccess: Training for
real-time intralingual subtitlers.

3
00:00:18,360 --> 00:00:21,800
This is unit 3, IT competence.

4
00:00:22,080 --> 00:00:27,480
In this presentation, we will be
dealing with element 1: Input tools.

5
00:00:29,200 --> 00:00:31,080
This will be an introduction

6
00:00:31,080 --> 00:00:36,240
to the input tools used by
real-time intralingual subtitlers.

7
00:00:36,840 --> 00:00:42,520
The materials for this presentation
have been created by Sub-Ti Access,

8
00:00:42,520 --> 00:00:46,000
an accessibility service provider
from Italy.

9
00:00:46,320 --> 00:00:48,800
My name is Enrico Pigliacampo,

10
00:00:49,160 --> 00:00:53,080
I am an Italian man
with a beard and short hair.

11
00:00:53,800 --> 00:00:59,120
I will be displayed in a box
at the top right corner of each slide.

12
00:01:01,200 --> 00:01:04,120
The learning outcome
for this presentation

13
00:01:04,400 --> 00:01:07,480
is to be able
to explain the differences,

14
00:01:07,480 --> 00:01:10,400
advantages and disadvantages

15
00:01:10,400 --> 00:01:13,760
of the different input tools
available on the market

16
00:01:13,760 --> 00:01:16,960
for both respeaking
and typing techniques.

17
00:01:19,000 --> 00:01:23,080
In this video presentation we will
be dealing with the following topics:

18
00:01:23,720 --> 00:01:28,720
Input tools for typing techniques,
input tools for respeaking

19
00:01:28,720 --> 00:01:31,440
and input tools for ASR,

20
00:01:31,440 --> 00:01:35,080
which is short for
Automatic Speech Recognition.

21
00:01:37,320 --> 00:01:42,920
The first topic we will talk about
is input tools for typing techniques.

22
00:01:43,520 --> 00:01:48,160
Among the different tools that
real-time intralingual subtitlers

23
00:01:48,160 --> 00:01:51,000
and speech-to-text reporters use,

24
00:01:51,200 --> 00:01:56,600
we have chosen two that are
the best-known ones plus Velotype,

25
00:01:56,600 --> 00:02:01,240
which is one of the specific
techniques of the LTA project.

26
00:02:03,440 --> 00:02:06,720
Some real-time
intralingual subtitlers

27
00:02:06,720 --> 00:02:10,400
use the PC keyboard
as an input tool.

28
00:02:10,400 --> 00:02:13,840
It is easy to find,
every computer has one,

29
00:02:13,840 --> 00:02:16,280
and they’re cheap
if you need to buy one.

30
00:02:16,840 --> 00:02:21,360
Although there are techniques
to write faster on the PC keyboard,

31
00:02:21,520 --> 00:02:24,200
it is difficult
to keep up with the speech

32
00:02:24,200 --> 00:02:27,920
as it was not created
for this specific purpose.

33
00:02:28,760 --> 00:02:32,400
The best-known PC keyboard is
the QWERTY keyboard.

34
00:02:33,000 --> 00:02:37,640
Although many other PC keyboards have
been developed for different uses,

35
00:02:38,040 --> 00:02:40,880
some even to increase typing speed,

36
00:02:40,880 --> 00:02:44,200
the QWERTY keyboard remains
the most commonly used

37
00:02:44,200 --> 00:02:45,960
and the most famous one.

38
00:02:48,800 --> 00:02:52,000
One of the most famous
typing techniques

39
00:02:52,000 --> 00:02:56,680
for real-time intralingual subtitling
and speech-to-text reporting

40
00:02:56,680 --> 00:02:58,200
is stenotype.

41
00:02:58,720 --> 00:03:02,880
Modern stenotype machines
resemble a small piano keyboard.

42
00:03:03,440 --> 00:03:06,920
Stenographers can
write texts very fast

43
00:03:06,920 --> 00:03:09,160
thanks to a combination of keys.

44
00:03:09,360 --> 00:03:15,360
An expert stenographer can reach
a speed up to 300 words per minute.

45
00:03:16,160 --> 00:03:20,880
Stenotype machines are quite expensive,
and the technique is hard to learn

46
00:03:20,880 --> 00:03:25,560
especially because in some countries
there are very few training courses.

47
00:03:27,760 --> 00:03:32,520
The Velotype keyboard was created
by a Dutch company called Velotype.

48
00:03:32,960 --> 00:03:37,600
It was created specifically
for real-time intralingual subtitlers

49
00:03:37,600 --> 00:03:39,680
and speech-to-text reporters.

50
00:03:40,160 --> 00:03:43,200
The keyboard enables you
to write very fast

51
00:03:43,200 --> 00:03:46,400
and keep up even with
the fastest speeches.

52
00:03:46,920 --> 00:03:49,880
By pressing several keys
at the same time,

53
00:03:49,880 --> 00:03:52,840
you can write syllables
and complete words.

54
00:03:53,160 --> 00:03:55,440
The technique is easy to learn.

55
00:03:55,440 --> 00:03:58,640
You can learn the basics
in a few months.

56
00:03:59,640 --> 00:04:02,560
Although you have to buy,
or lease, a new keyboard,

57
00:04:02,760 --> 00:04:07,680
it is cheaper than other machines
and more efficient than a PC keyboard.

58
00:04:09,320 --> 00:04:11,640
In the second section
of this presentation,

59
00:04:11,640 --> 00:04:14,800
we will talk about
input tools for respeakers.

60
00:04:15,920 --> 00:04:19,800
Respeakers can choose
from among many different products

61
00:04:19,800 --> 00:04:21,480
for each tool they use.

62
00:04:21,840 --> 00:04:25,200
The specific tools
and software each respeaker use

63
00:04:25,200 --> 00:04:30,040
should be chosen according to
the context in which he or she works.

64
00:04:30,560 --> 00:04:33,360
Of course the main tool
is the computer.

65
00:04:33,520 --> 00:04:37,240
Each respeaker must choose
the most appropriate PC

66
00:04:37,240 --> 00:04:40,680
depending on his
or her professional needs.

67
00:04:43,120 --> 00:04:45,160
As well as a computer,

68
00:04:45,160 --> 00:04:49,480
the main tools a respeaker needs
are headphones and a microphone.

69
00:04:50,200 --> 00:04:53,680
Headphones are useful
to clearly hear the speaker.

70
00:04:54,080 --> 00:04:57,360
Professionals may opt
not to use headphones

71
00:04:57,360 --> 00:05:01,600
if they are working in
a soundproof booth with a loudspeaker.

72
00:05:02,080 --> 00:05:05,200
They may also decide
not to use headphones

73
00:05:05,200 --> 00:05:08,120
if they are in the same room
as the speaker.

74
00:05:08,520 --> 00:05:11,560
However, if there are
many people in the room,

75
00:05:11,560 --> 00:05:16,360
it is better to have the audio input
from the microphone into the headphones,

76
00:05:16,760 --> 00:05:19,720
so as to cancel out
background noises.

77
00:05:20,440 --> 00:05:22,840
The microphone
is a very important tool.

78
00:05:23,000 --> 00:05:25,200
It is connected to the computer

79
00:05:25,200 --> 00:05:29,920
and the respeaker dictates the text
to a speech recognition software

80
00:05:29,920 --> 00:05:31,480
through the microphone.

81
00:05:32,040 --> 00:05:36,000
If the respeaker is working
in a crowded room with a lot of noise,

82
00:05:36,000 --> 00:05:38,360
the stenomask is the perfect tool.

83
00:05:38,800 --> 00:05:42,800
It is a microphone built into
a padded, soundproof enclosure.

84
00:05:43,160 --> 00:05:47,360
This ensures acoustic isolation
from external noise

85
00:05:47,360 --> 00:05:50,800
and the microphone captures
only the respeaker’s voice.

86
00:05:52,840 --> 00:05:56,560
Every respeaker needs
a speech recognition software.

87
00:05:57,400 --> 00:06:00,920
It is a software that turns
the vocal input from the microphone

88
00:06:00,920 --> 00:06:02,680
into written text.

89
00:06:03,040 --> 00:06:08,120
The transcript can then be visualised
on other software programmes

90
00:06:08,120 --> 00:06:11,000
which the professional
uses to edit it

91
00:06:11,000 --> 00:06:13,600
and then broadcast
it to end-users.

92
00:06:14,240 --> 00:06:17,400
We will talk about
this kind of software

93
00:06:17,400 --> 00:06:21,320
in the lectures on editing software
and output tools.

94
00:06:22,600 --> 00:06:26,320
The most famous speech recognition
among respeakers

95
00:06:26,320 --> 00:06:30,760
is Dragon Naturally Speaking,
but it is not the only one.

96
00:06:30,760 --> 00:06:33,520
Newton is also used by professionals.

97
00:06:33,760 --> 00:06:36,800
The PerVoice Subtitling Workstation

98
00:06:36,800 --> 00:06:40,280
also uses speech recognition
to create subtitles.

99
00:06:40,680 --> 00:06:44,240
There are also software programmes
which are open source

100
00:06:44,240 --> 00:06:46,080
and free of charge.

101
00:06:46,080 --> 00:06:49,160
You can search online
and try them out

102
00:06:49,160 --> 00:06:52,280
to see how a speech recognition
software works.

103
00:06:54,360 --> 00:06:58,480
ASR stands
for Automatic Speech Recognition.

104
00:06:58,880 --> 00:07:01,240
In the last section of
this presentation,

105
00:07:01,240 --> 00:07:04,320
we will talk very briefly
about this kind of software.

106
00:07:05,440 --> 00:07:09,960
Automatic speech recognition software
recognises any speaker’s speech

107
00:07:09,960 --> 00:07:11,920
and transcribes it.

108
00:07:12,240 --> 00:07:14,040
With this kind of software

109
00:07:14,040 --> 00:07:17,560
there‘s no need of a respeaker
to repeat the text.

110
00:07:18,280 --> 00:07:22,520
Of course, the software will
transcribe everything the speaker says

111
00:07:22,520 --> 00:07:26,720
and make mistakes when it does not
recognise the right words.

112
00:07:26,720 --> 00:07:31,840
That is why there must always be
a text editor to monitor the transcript,

113
00:07:31,840 --> 00:07:34,440
spot any mistakes
and correct them.

114
00:07:36,360 --> 00:07:37,600
Summary.

115
00:07:39,080 --> 00:07:43,080
To sum up, there are two main points
to keep in mind

116
00:07:43,080 --> 00:07:45,880
when choosing
your tools and techniques.

117
00:07:46,400 --> 00:07:50,120
Firstly, in order to choose
the best technology,

118
00:07:50,120 --> 00:07:54,880
you need to know the work environment
in which you are going to work.

119
00:07:55,640 --> 00:07:59,320
It is also good to try
different techniques and tools

120
00:07:59,320 --> 00:08:02,320
to understand which one
suits you best

121
00:08:02,320 --> 00:08:06,720
and will help you to produce
better real-time intralingual subtitles.

122
00:08:08,080 --> 00:08:09,720
Exercises.

123
00:08:11,120 --> 00:08:17,000
Search online for at least 3 free
speech recognition software applications

124
00:08:17,000 --> 00:08:18,760
and try them out.

125
00:08:19,240 --> 00:08:23,320
At the end of the exercise,
select the best one for you.

126
00:08:25,040 --> 00:08:32,400
[LTA MUSIC]

127
00:08:35,640 --> 00:08:38,120
LTA, Live Text Access.

128
00:08:38,760 --> 00:08:41,640
Universitat Autònoma de Barcelona.

129
00:08:42,320 --> 00:08:45,920
SDI, Internationale Hochschule.

130
00:08:46,600 --> 00:08:50,560
Scuola Superiore
per Mediatori Linguistici.

131
00:08:51,360 --> 00:08:53,200
ZDF Digital.

132
00:08:54,040 --> 00:08:57,720
European Federation
of Hard of Hearing People, EFHOH.

133
00:08:58,480 --> 00:09:00,000
Velotype.

134
00:09:00,360 --> 00:09:02,120
Sub-Ti Access.

135
00:09:02,840 --> 00:09:08,080
European Certification and Qualification
Association, ECQA.

136
00:09:11,280 --> 00:09:15,480
Co-funded by the Erasmus+ Programme
of the European Union.

137
00:09:17,120 --> 00:09:19,120
Erasmus+ Project:

138
00:09:19,360 --> 00:09:31,520
2018-1-DE01-KA203-004218

139
00:09:32,440 --> 00:09:36,080
The information and views
set on this presentation

140
00:09:36,200 --> 00:09:38,000
are those of the authors

141
00:09:38,000 --> 00:09:41,760
and do not necessarily reflect
the official opinion

142
00:09:41,760 --> 00:09:43,640
of the European Union.

143
00:09:44,440 --> 00:09:48,160
Neither the European Union
Institutions and bodies

144
00:09:48,520 --> 00:09:51,480
nor any person
acting on their behalf

145
00:09:51,840 --> 00:09:54,640
may be held responsible
for the use

146
00:09:54,880 --> 00:09:58,600
which may be made
of the information contained here.