Guidage Gestuel pour des Robots Mobiles

Abstract : Using a visuo-tactil interface may be restrictive when mobility and situation awareness are required. This is particularly problematic in hostile environment as commanding a drone on a battlefield.In the work presented here, we hypothesize that gesture is a less restrictive modaility as it doesn't require to manipulate nor to look at a device.Thus we followed a user-centered approach to confirm practical advantages and technical feasibility of gestural interaction for drones.First, the theoretical study of gestures allowed us to design an interaction model. It consists on activating commands by executing standardized semaphoric gestures. Sound messages inform the user and a confirmation mechanism secure the interaction.Second, a gestural vocabulary has been built. To do so, a methodology has been proposed and used : end users elicited then elected the most appropriate gestures.Then, the interaction model and the selected gestures have been evaluated. A laboratory study showed that they are both easy to learn and use and helps situation awareness.An interactive system as then been developed. It's architecture has been deducted from our interaction model and a gesture recognizer as been buit. Different from usual methods, the recognizer we proposed is based on formal description of gestures using regular expressions.Finaly, we conducted a user testing of the proposed system. The evaluation by end-users confirmed our hypothesis that gestures are a modality less distractive both visualy and physicaly.
Keywords : Gestures Hmi
Document type :
Theses
Complete list of metadatas

Cited literature [20 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/tel-01814263
Contributor : Abes Star <>
Submitted on : Wednesday, June 13, 2018 - 10:19:34 AM
Last modification on : Tuesday, May 21, 2019 - 5:41:13 PM
Long-term archiving on: Friday, September 14, 2018 - 7:55:32 PM

File

2016PSLEM096_archivage.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01814263, version 1

Citation

Florent Taralle. Guidage Gestuel pour des Robots Mobiles. Interface homme-machine [cs.HC]. PSL Research University, 2016. Français. ⟨NNT : 2016PSLEM096⟩. ⟨tel-01814263⟩

Share

Metrics

Record views

228

Files downloads

1019