Grounding Language Understanding GLU2017

August 25, 2017, KTH Royal Institute of Technology, Stockholm, Sweden

Workshop Venue: Salongen, KTH Library

Osquars backe 31, 114 28 Stockholm, Sweden

Closest underground (T-bana) stop: Tekniska Högskolan (red line towards Mörby Centrum)

Lunch: Syster o Bror

Drottning Kristinas väg 24, 114 28 Stockholm

Dinner: Stora Skuggans Wärdshus

Stora Skuggans Väg 12, 115 42 Stockholm, Sweden

Closest underground (T-bana) stop: Universitetet (red line towards Mörby Centrum), 15 minutes walk

Walking there directly from the workshop venue takes around 30 minutes

Presentation instructions

Oral presentaions: please prepare an oral presentation of 15 minutes plus 5 minutes for questions. Given the tight schedule in the programme, it is essential to keep within those time limits, please make sure you do not exceede them. You may use your own laptop.

Poster presentations: please prepare a poster in A0 format. Best if portrait orientation. You will have two slots of 30 minutes each (coffee breakes) to present your work. Please make sure your poster is in place before the beginning of the workshop. We will provide means to hang the poster.

Keynote presentations: Keynote speakers have 40 minutes for their presentations, including questions. It is up to each talker to decide how to divede this time into presentation and questions.

If you do not know what presentation form you have been assigned, please check the Programme page.

Paper submission

We invite original submissions of three to five pages papers in INTERSPEECH format. Each submission will be reviewed by at least two members of the Scientific Review Committee.

Submit your paper through EasyChair: https://easychair.org/conferences/?conf=glu2017

Submission deadline: Wednesday, 31 May 2017 Sunday, 4 June 2017, 23:59 CET

Paper submissions will be reviewed by at least two specialists of the area to which it belongs.

Camera-ready submission

Authors of accepted papers are required to submit a reivsed version of their paper through the EasyChair url above before Wednesday, 26 July 2017. Please include full author and affiliation information in the final version of the paper. Follow the guidelines below.

Paper preparation guidelines

The format is the same as for INTERSPEECH 2017. Accepted papers will have a minimum of three pages including references and a maximum of 4 pages of text, a fifth page is allowed for references only.

The Author’s Kit that comprises of templates for LaTeX and MSWORD and can be downloaded here

It is the author’s responsibility to ensure that submitted papers use the correct styles, fonts, etc. Paper submissions must be in PDF format. The PDF file should comply with the following requirements:

All authors must proofread their PDF file prior to submission to ensure it is ready for publication if the paper is accepted. Authors should not rely on proof-reading the source document – please proofread the PDF file. The proceedings editors will contact authors of non-complying papers to obtain a replacement, to be submitted by a specified date. To ensure that there is no delay to conference proceedings, papers for which a replacement is not provided by the specified date, will be excluded from the proceedings.

Programme

Scientific programme for the workshop on Friday, August 25.

The full proceedings with foreword and indexes will be available shortly, you can click on the links to download each single article/abstract.

TimeDescription
Getting name badges, hanging posters, uploading presentations, ...
Welcome and Introduction
Keynote 1: Binding through assemblies in the human brain: Recent data and models [abstract],
Robert Legenstein
Streching legs
Relational Symbol Grounding through Affordance Learning: An Overview of the ReGround Project [pdf],
Laura Antanas, Jesse Davis, Luc De Raedt, Amy Loutfi, Andreas Persson, Alessandro Saffiotti, Deniz Yuret, Ozan Arkan Can, Emre Unal and Pedro Zuidberg Dos Martires
Communication with Speech and Gestures: Applications of Recurrent Neural Networks to Robot Language Learning [pdf],
Alexandre Antunes, Gabriella Pizzuto and Angelo Cangelosi
Poster Session 1 (Coffee break)

Automatic Speaker’s Role Classification with a Bottom-up Acoustic Feature Selection [pdf],
Vered Silber-Varod, Anat Lerner and Oliver Jokisch

Comparison of Effect of Speaker's Eye Gaze on Selection of Next Speaker between Native- and Second-Language Conversations [pdf],
Koki Ijuin, Takato Yamashita, Tsuneo Kato and Seiichi Yamamoto

Grounding Imperatives to Actions is Not Enough: A Challenge for Grounded NLU for Robots from Human-Human data [pdf],
Julian Hough, Sina Zarriess and David Schlangen

Interactive Robot Learning of Gestures, Language and Affordances [pdf],
Giovanni Saponaro, Lorenzo Jamone, Alexandre Bernardino and Giampiero Salvi
Proposal of a Generative Model of Event-based Representations for Grounded Language Understanding [pdf],
Simon Brodeur, Luca Celotti and Jean Rouat
Vision-based Active Speaker Detection in Multiparty Interaction [pdf],
Kalin Stefanov, Jonas Beskow and Giampiero Salvi
Finding Regions of Interest from Multimodal Human-Robot Interactions [pdf],
Pablo Azagra, Javier Civera and Ana C. Murillo
Lunch
Keynote 2: Recursion all the way: in Language, Action and Semantic Association [abstract],
Katerina Pastra
Building a multimodal lexicon: Lessons from infants' learning of body part words [pdf],
Rana Abu-Zhaya, Amanda Seidl, Ruth Tincoff and Alejandrina Cristia
SPEECH-COCO: 600k Visually Grounded Spoken Captions Aligned to MSCOCO Data Set [pdf],
William Havard, Laurent Besacier and Olivier Rosec
Visually Grounded Word Embeddings and Richer Visual Features for Improving Multimodal Neural Machine Translation [pdf],
Jean-Benoit Delbrouck, Stéphane Dupont and Omar Seddati
Poster Session 2 (Coffee break)

Sparse Autoencoder Based Semi-Supervised Learning for Phone Classification with Limited Annotations [pdf],
Akash Kumar Dhaka and Giampiero Salvi

Towards a Knowledge Graph based Speech Interface [pdf],
Ashwini Jaya Kumar, Sören Auer, Christoph Schmidt and Joachim Köhler

Enhancing reference resolution in dialogue using participant feedback [pdf],
Todd Shore and Gabriel Skantze

Keynote 3: Towards Autonomous Language Learning [abstract]
Emmanuel Dupoux
Language is Not About Language: Towards Formalizing the Role of Extra-Linguistic Factors in Human and Machine Language Acquisition and Communication [pdf],
Okko Räsänen
Analysis of Audio-Visual Features for Unsupervised Speech Recognition [pdf],
Jennifer Drexler and James Glass
Partitioning of Posteriorgrams using Siamese Models for Unsupervised Acoustic Modelling [pdf],
Arvid Fahlström Myrman and Giampiero Salvi
Closing Remarks
Dinner

Registration

We are trying to get the costs as low as possible.

Registration fees will be as follows:

to all fees you need to add 25% VAT according to Swedish laws.

The corresponding fees in Swedish crowns, listed on the Intespeech registration page are

the above fees include the VAT.

Registration and payment are handled through the Interspeech registration service.

If you wish to register to the workshop but not to the main conference, click on "satellite workshops only".

If you wish to do an early bird registration to the main conference, but do not know if you want to register to the workshop at that point in time, you can always register to the workshop later (but always before the workshop registration deadline on August 2).

Organizers

The workshop is organized by

And supported by the CHIST-ERA IGLU consortium.

For information, contact Giampiero Salvi.

Registration

We are trying to get the costs as low as possible.

Registration fees will be as follows:

all fees are excluding VAT.

Registration and payment are handled through the Interspeech registration service.

If you wish to register to the workshop but not to the main conference, click on "satellite workshops only".

If you wish to do an early bird registration to the main conference, but do not know if you want to register to the workshop at that point in time, you can always register to the workshop later (but always before the workshop registration deadline on August 2).

Scientific Committee

International Workshop on Grounding Language Understanding, Satellite of Interspeech 2017

The purpose of this workshop is to bring together researchers working with modeling different aspects of grounding language acquisition and understanding.

Although speech and language technology have reached a level of maturity, machines still fall short of human performance, especially when considering flexibility and robustness. It is therefore desirable to extend the machine learning approach which has been applied to speech and language technology, and emulate more closely the way humans and living systems acquire language. Desirable properties of the new learning approaches would include:

This requires a highly multidisciplinary approach, combining the field of human language processing, speech technology, machine learning, developmental robotics and cognitive and systems neuroscience.

Confirmed Keynote Speakers

Scientific Area Topics

Important Dates

Submission deadline:
Wednesday, 31 May 2017
Sunday, 4 June 2017, 23:59 CET.

Paper Notification Acceptance:
Wednesday, 28 June 2017.

Publication of Full Programme:
Wednesday, 12 July 2017.

Camera-ready Paper Due:
Wednesday, 26 July 2017.

Registration Deadline:
Friday, 4 August 2017.

Workshop date: Friday, 25 August 2017.

Endorsement

Official INTERSPEECH 2017 Satellite Event

Supported by the International Speech Communication Association

International Speech Communication Association

Supported by the CHIST-ERA Network

CHIS-ERA Network