• Aucun résultat trouvé

A speech translation system for medical dialogue in sign language — Questionnaire on user perspective of videos and the use of Avatar

N/A
N/A
Protected

Academic year: 2022

Partager "A speech translation system for medical dialogue in sign language — Questionnaire on user perspective of videos and the use of Avatar"

Copied!
4
0
0

Texte intégral

(1)

Proceedings Chapter

Reference

A speech translation system for medical dialogue in sign language — Questionnaire on user perspective of videos and the use of Avatar

Technology

BOUILLON, Pierrette, et al .

Abstract

This paper summarizes the findings of a questionnaire conducted with the Deaf community infrancophone Europe. It aimed to gather feedback on specific features to be included in the sign language videos of BabelDr, a medical speech-to-sign translation platform. The results show that the whole sentence should be subtitled, the signer should be shown front-o n only, and the background should be of a light colour. Explanatory images can be added to clarify the medical content. Lastly, human videos are preferred to avatars in this specific context, even if more than a third of the respondents think that an avatar could be useful in this context.

BOUILLON, Pierrette, et al . A speech translation system for medical dialogue in sign language

— Questionnaire on user perspective of videos and the use of Avatar Technology. In:

Proceedings of the 3rd Swiss Conference on Barrier-free Communication (BfC 2020) . Winterthur : ZHAW Zürcher Hochschule für Angewandte Wissenschaften, 2021. p. 46-54

Available at:

http://archive-ouverte.unige.ch/unige:151996

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

A Speech Translation System for Medical Dialogue in Sign Language - Questionnaire on User Perspective of Videos and the

Use of Avatar Technology

Pierrette Bouillon Bastien David, Irene Strasly

Faculté de Traduction et d’Interprétation Genève – CH

pierrette.bouillon, bastien.david, irene.strasly

@unige.ch

Hervé Spechbach Geneva University Hospitals

Genève – CH hervé.spechbach

@hcuge.ch

Appendix. Questionnaire 1. Respondents

a) What sign language do you use?

i. LSF ii. LSFB iii. LSF-SR b) Are you 18 years old or over?

i. Yes i. No

2. Background of the Video

a) Do you consider that the kind of background behind the avatar is important?

i. Yes ii. No

b) Which kind of background would you choose?

i. ii. iii.

c) What kind of color do you prefer?

i. ii. iii.

iv. v. vi. Other d) What image would you prefer?

i. ii. iii.

iv. v. vi.

e) Do you have anything to add?

3. Additional Image

a) Do you consider that the image behind the avatar could give useful information?

i. Yes ii. No

b) When referring to an organ or a certain kind of pain, which image do you prefer?

i. ii. iii.

c) Which image is the most understandable?

i. ii. iii.

(3)

d) What colour do you prefer?

i. ii. iii.

e) Where would you place the avatar in the video?

i. ii. iii.

f) Do you have anything to add?

4. Subtitles

a) Do you think that subtitles provide useful information?

i. Yes ii. No

b) What kind of subtitles would you prefer?

i. One word only ii. The complete sentence iii. One letter at a time

c) Where would you like the subtitles?

i. Below ii. On the side

d) What color would you like the subtitles to be?

i. White over black ii. Black over white iii. White over blue iv. Black over red v. Transparent black vi. Transparent white

e) To represent a word that does not have a sign (for example: Dafalgan), you prefer ...

i. A subtitle ii. An image

iii. The interpreter signs using the manual alphabet f) Do you have anything to add?

5. Screen Format and Size

a) Do you think that the size of the screen and the size of the avatar are important?

i. Yes ii. No

b) Which screen orientation do you prefer?

i. Portrait ii. Landscape c) What kind of screen would you prefer?

i. ii. iii.

iv. v. vi.

d) Which image would you choose?

i. ii. iii.

e) What is the ideal distance between the head of the avatar and the top of the screen?

i. ii. iii.

f) Do you have anything to add?

6. Advertising

a) What advertising information would you agree to see on the videos?

i. A logo ii. Our partners

(4)

iii. Our sponsors iv. No advertisement b) Where would you put this information?

i. Before the video ii. During the video iii. After the video iv. Behind the avatar c) Do you have anything to add?

7. Perception of Three-Dimensional Avatars

a) Do you consider that videos with avatars can be useful?

i. Yes ii. No

b) In this medical context, would you prefer videos with filmed humans or virtual characters?

i. Filmed humans ii. Virtual characters

c) To better understand the signer, which video would you prefer?

i. ii.

iii. iv. v.

d) Do you have anything to add?

Références

Documents relatifs

A speech translation system for medical dialogue in sign language — Questionnaire on user perspective of videos and the use of Avatar Technology..

The hybrid approach combines the best neural ma- chine translation model with the best classification model to build an N-best list of sentences, in this experiment a 2-best list

BabelDr is a medical speech to speech translator, where the doctor has to approve the sentence that will be translated for the patient before translation; this step is done

While most aspects unable to assess whether left hemisphere dominance of spoken languages rely on fast acoustic transitions arises from the analysis of the

The experimental version of the system uses a hybrid ap- proach to speech understanding which combines the current grammar-based processing route with robust processing based on

The analysis of the elicited queries points to a number of system requirements, both from the meeting recording and processing, and from the retrieval perspectives.. While not all

The lexicon is actually possible to write manually (Morphological tools like Mmorph (Petitpierre/Russell, 1995), or PC- Kimmo (Koskenniemi, 1983) are not integrated in

The help module’s function in this application is to accelerate the learning effect. By providing the user with a list of sentences related to the one that has been