Dear all,

* Apologies for cross-posting *

Speech interfaces are customary in many types of robots and robotic applications. Despite the progress in speech recognition and many other areas of natural language processing in recent years, failures of speech interfaces in robotic scenarios are numerous and commonplace. In contrast to the shared experience of failure of speech interfaces in robotics, the literature is positively skewed towards the success and good performance of these. The documentation of failure speech interfaces on the other hand is exceedingly rare and a systematic study of failure and its causes is non-existent.

The WTF workshop aims at bringing together a multidisciplinary group of researchers from the fields of robotics, human-Robot Interaction (HRI), natural language processing, conversation analysis and pragmatics and industrial partners to address this gap. The workshop aims to provide a platform to discuss the multitude of failures of speech interfaces openly and, if possible, systematically, in the hope of identifying the most fruitful directions for overcoming these failures in future systems. The workshop will focus on human-robot joint action scenarios involving multimodal coordination between humans and robots as these are the norm in scenarios where robotic speech interfaces are deployed.

Format:

The workshop will take place in two parts, with the first part being online, and the second part being on site at the University of Hertfordshire. The two parts are complementary, and the work of the on-site part will build on the outcome from the online part.

Both parts will consist of a combination of keynote talks by experts in the field followed by bash talks and group discussions on the different types of failures.

Workshop website:

https://sites.google.com/view/wtfworkshop2022/

IMPORTANT: While the online part was held on the 16^th and 17^th of June, we are now reopening the call for participation for the on site part of the 5^th of September. Interested participants who did not take part in the online session are welcome to apply and join us in the discussion!

Dates:

• Application: 22^nd July 2022 (Notification of Acceptance on the 29^th)

• Workshop - On-site part: 5^th September 2022, full day (9:30am to 5pm)

Applications for participation should be submitted before or on the 22^nd of July via the following link:

https://sites.google.com/view/wtfworkshop2022/dates-and-submission

We have limited funding available to cover travel and accomodation costs for travel within the UK. See the workshop website for details.

Outcomes:

Apart from hopefully lively discussions and exchanges about the most pressing failures in such speech interfaces, the intended outputs of this workshop are twofold:

• A position paper summarizing the kinds of failures commonly observed by researchers working on (or with) robotic speech interfaces, and the most promising future directions to overcome these,

• The identification of a prototypical application scenario embodying most of the identified challenges that may act as a future benchmark.

Organizers:

• Frank Förster, University of Hertfordshire

• Christian Dondrup, Heriot-Watt University

• Joel Fischer, University of Nottingham

• Marta Romeo, University of Manchester

• Patrick Holthaus, University of Hertfordshire

• Luke Wood, University of Hertfordshire

Contact:

For all inquiries or early expressions of interest, contact Frank Foerster (f.foerster@herts.ac.uk)