Extended Multimodal Search and Communication Systems for Safe In-Car Application
The aim of the proposed project is to develop a system for safe information access and communication while driving.
The aim of the proposed project is to develop a system for safe information access (search, navigation, point of interest) and communication (texting) while driving. In order to reach that goal, we approach the problem from a holistic view, investigating the underlying driving forces, studying the goals underlying searching and texting. Special attention will be paid to task and context factors such as multi-tasking and the associated cognitive load for the driver. Evaluations will be conducted in cycles that will be synchronized with the prototype development. In particular, the project seeks to use/develop methods and metrics for driving assessment that can be used at all stages of system development.
GetHomeSafe will go beyond the state of the art by:
Additionally, the objectives of the proposed project include providing an open source toolbox with an open source driving simulation software as a centrepiece.
- Driving further the introduction of industry standards for supporting wireless communication between car and mobile devices
- Not only targeting one specific communication goal of drivers but comprehensively investigating the entirety of ÃÂ¢Ã¯Â¿Â½Ã¯Â¿Â½driving forcesÃÂ¢Ã¯Â¿Â½Ã¯Â¿Â½
- Introducing reliable and easy-to-use speech recognition (ASR) for information access (search) and messaging
- Supporting multimodal text input and access to information sources such as searching the web for a driving destination or information about it
- Supporting specific interaction goals directly by circumventing explicit interaction (i.e. search, composing of text messages)
Group: Speech Communication and Technology
Joakim Gustafson (Project leader)
Duration: 2011-12 - 2014-12
Proceedings of the EACL Satellite Workshop Dialogue In Motion (DIM-2014). Gothenburg, Sweden. [abstract] [pdf] (2014). Human pause and resume behaviours for unobtrusive humanlike in-car spoken dialogue systems. In
Proc. of Multimodal Corpora 2013. Edinburgh, UK. [abstract] [pdf] (2013). Temporal precision and reliability of audience response system based annotation. In
In Proc. of Workshop on Innovation and Applications in Speech Technology (IAST). Dublin, Ireland. [abstract] [pdf] (2012). Investigating negotiation for load-time in the GetHomeSafe project. In