ALADIN

This page shows some video's of the ALADIN project. The goal of this project is to develop a speech interface which can learn directly from user interaction. It is aimed at speech impaired users, for which conventional speech interfaces may not perform adequately due to their deviant (and typically dialected) speech.

ALADIN iMinds video

The material for this video was recorded at the iMinds conference 2013. In the video, you can see the demonstrator operating home automation devices in a 3D home environment using spoken commands. Additionally, a tablet interface provides feedback, allows manual control and will assist in the training process. The quoted website aladinspeech.be is currently under construction.

Video of the training process used for the ALADIN demo

The ALADIN vocal interface is currently trained prior to usage in a very short session. In the final system, the system will continuously train from user feedback, thus eliminating a seperate training stage.

Video of a prospective ALADIN user

The user can be seen operating the lights in a 3D model of a home environment, using a tablet interface and using speech control. The user has impaired speech and difficulties with manual operating devices.