音声ブラウザご使用の方向け: SKIP NAVI GOTO NAVI

Web Posted on: August 4, 1998


Prospects for enhanced captioning services for the hearing impaired

 

P.A.M.E. Crombag-Röben
L. Groeneweg-Bruckman
M. Partouns
M. Soede

Institute for Rehabilitation Research
PO box 192, NL-6430 AD Hoensbroek
The Netherlands
tel: +31.45.5237537
fax: +31.45.5231550
email: p.crombag@irv.nl
msoede@irv.nl
University of Maastricht/CWS
Maastricht, The Netherlands


1. Summary:

This paper presents the first part of the work of the Caption Life project. The project aims for a fast and easy accessible relay service where spoken text is simultaneously transformed in typed text on a screen. The focus is on the critical factors to be addressed in the research and development as there are the training of the velotypists, aspects of the role of the captioning service in communication of the hearing impaired, and an adapted text telephone suitable to use the relay service with Velotype text input speed.



| Top |

2. Introduction and goal of the project:

The project Caption Life uses a fast text input system: the Velotype system. With this text input system a (velo)typist can produce text at full speaking speed. It can be used in "normal" captioning services but also new relay services will be developed like a life captioning in a text telephone tele-relay situation.
The goal of this development is to give a faster relay service than ever: speech-to-text conversion with a speed of normal speech, and a much more flexible usage of the service. Via a tele-relay situation an almost immediate availability of the speech-to-text relay service is foreseen. An underlying goal for the project is the training of the veloptypists. The Velotype keyboard is completely different from the Dvorak- and Qwerty keyboards, and the typist has to be trained to obtain a sufficient speed with which speech spoken with "normal" speed can be followed.
During the project two test sites will be used to try out the velotype relay system and to evaluate this new concept of relay service. The test sites are located in The Netherlands and in Finland.

3. Project presentation:

3.1 Technical description

In figure 1 below, a typical situation is depicted for captioning speech of a normal speaking hearing person for the benefit of an hearing impaired person in a group discussion situation or a particular communication situation like the visit of a doctor.



Figure 1: principle model of the text telephone relay situation



3.2. Research

In the first year of the Caption Life project requirements of the users of text telephone relay services and of the organisation are collected via literature studies and direct contact with users having interviews and filling in questionnaires. Furthermore the organisation managers of the Dutch and Finnish relay services were interviewed as well as skilled velotypists.


3.2.1 The users of the relay service

Interviews were held with a small group of relay service users of the present text telephone relay services in The Netherlands and in Finland. The hearing impaired persons as well as the hearing communication partners were interviewed to get a good view on the critical factors of the present text telephone relay services and the user needs concerning the usage of the relay services and text telephones.

The following observations are made:

  • The initiative users are mostly the hearing impaired persons, of which only a small group uses the relay service on regular basis of at least several times (>10) per month.
  • Most conversations have a duration of about 5 to 15 minutes.
  • Speaking hearing impaired person do rather like to speak directly to the conversation partner (a hearing person), though most users are non-speaking hearing impaired persons.
  • The use of the relay centres is both for "business" like conversations as for social/private contacts. Most official business-like conversations are held during daytime and are quite short. Most social contacts are maintained in the evening hours.
  • In The Netherlands costs are a strong limiting factor to use the relay service frequently. This is because the Dutch relay service is not financially supported in any means, which is different from Finland and other European countries amongst others Sweden, where the users do not have to pay extra costs to use the relay service. The hearing person, using the relay service much less, is not worried about the costs.
  • Waiting periods to get in contact with the relay service and to get connected with the conversation partner are not experienced as too long by both the hearing users as the hearing impaired users, although sometimes it takes more than 30 seconds to get connected and be able to start the conversation.
  • The speed of the conversion from speech to text is experienced as reasonable or too slow. It takes about 10 to 30 seconds before getting the answer or reaction of the conversation partner.
  • Often the users have to adapt their speaking speed towards the typing speed of the operator.
  • Furthermore very often the conversion is not done literally by the operator because of the high speed of speech and the limited possibility of conversion by the text input system. The users prefer a literal conversion, including short expression made like eh, aha, oh yes, etc.

Confidentiality is a limiting factor in the usage of relay centres by hearing (impaired) users in both countries. The users mostly are aware of the third person listening to their conversation. Though all relay service do guarantee privacy of conversations (unfortunately our own experience did not confirm this guarantee).


3.2.2 The operators/velotypists of the relay service

On of the other goals of the Caption Life project is training new velotypists as the availability of skilled velotypists is little both in The Netherlands as in Finland. For the test sites at least ten trained and thus skilled velotypists are needed. For training the velotypists for the specific function of velotypist and operator as well a new velotype tutor programme, special dedicated to the specific tasks of the operator/velotypist, will be developed. The training of the operator/velotypist will also include a special training concerning deafness awareness.
The velotypists need at least a typing speed of 700 characters per minute to be able to type at "normal" speaking speed. To get this specific necessary skill it takes about 5 to 6 months training and practising.


3.2.3 The text telephone

A new text telephone will be designed and a initial series will be produced during the project. This is necessary because of the nowadays limited possibilities of international modem connections, and the limited possibilities of text telephones to receive a high speed of data and text. Most equipment is limited to about 110 bits/s, and this is too slow for receiving data from the Velotype text input system. Therefore a V18 bis ITU recommendation will be used. This recommendation will also make full-duplex communication possible to transmit simultaneously voice and text at the same time.



| Top |

4 Future project development and discussion

The project is now in the phase that the text telephone module is under a redesign and the Velotype tutor is in development. At the same time the training of the Velotypists will be organised. In The Netherlands the system will be build up as a so-called Call-centre. It will even be tested if the Velotypist can be made available in a tele-work situation. The evaluation methodology of the service is planned according the principles found in the Userfit handbook; effectiveness, efficiency and acceptability, measured in the experimental situation in Finland and The Netherlands. An alternative technology for captioning speech and conveying into text, that is often mentioned, is speech recognition. However continuous, speaker independent, and situation independent recognition of speech will not be available within the coming decennium. If so, it would be possible to replace the velotype and velotypist by such a speech recognition system. At the moment a first estimation about future usage of the velotype relay system is made via the interviews with the relay service users. Although the users had a positive attitude towards this new development and the consequences form them, it seems not quite possible to predict already the future usage of the velotype relay system. Increased usage is expected but is not clear how much.



| Top |

Acknowledgement:

The Caption Life consortium exists of the following partners: Velotype International/The Netherlands, IvD/The Netherlands, iRv/The Netherlands, Comterm/Sweden, NovaCall/Finland and Centre Electre/France.



| Top |

Literature:

Use of telecommunication: the needs of people with disabilities, Frederiksen, J. ea. (eds.), Madrid,.1989 (p. 32-35)

Issues in telecommunication and disability, Von Tetzchner, S. (eds.), ECSC-EEC-EAEC, Brussels*Luxembourg, CD-NA-13845-EN-C, 1991

Proceedings of the COST 219 conference on Telecommunications and people with disabilities - current practice and future plans in Europe, Astbrink, G.(eds.), ECSC-EEC-EAEC, Brussels*Luxembourg, 1992

Telecommunications for all, P.R.W. Roe (eds.), ECSC-EC-EAEC, Brussels*Luxembourg, CD-90-95-712-EN-C, 1995

Survey of text telephones and relay services in Europe, (eds.) K.G. Olesen, DG XIII, EUR 14242, 1992

Döpping, O. The chord keyboard report writing clear text with the speed of speech, Swedish Handicap Institute, Sweden, 1988

European service delivery systems in rehabilitation technology, HEART line C, iRv, The Netherlands, Drukkerskollektief Geulle, 1994

Rapportage Gehandicapten, J.M. Timmermans, SCP, Rijswijk, The Netherlands, 1994

Handboek WVG, Handboek AWBZ, Handboek AAW, 1996

TNO-PG rapport: Slechthorendheid in Nederland, A.M.J. Chorus ea., Leiden, The Netherlands, 1995

Boxsel, J.A.M., Ch.G. Willems. Kansen en benaderingen voor Assistive Technology, Rathenau Instituut, Den Haag, 1997; S32

Userfit, a practical handbook on user-centred design for AT, TIDE-User consortium, ed. Poulson et all, ECSC-EC-EAEC, Brussels*Luxembourg, 1996



| Top | | TIDE 98 Papers |