WWW 2018 Challenge: Semantic Visual Question Answering

Question Answering Mediated by Visual Clues and Knowledge Graphs

WWW 2018 LYON, FRANCE (23 - 27 April 2018)

Website: https://visual-question-answering-challenge.github.io/

Summary:

This challenge focuses on the use of semantic representation methods to support Visual Question Answering. Given a large image collection, find a set of images matching natural language queries. The task will support advancing the state-of-the-art in Visual Question Answering by focusing on methods which explore the interplay between contemporary machine learning techniques, semantic representation and reasoning mechanisms.
Topics of interest for the challenge include (but are not restricted to):

- Visual Question Answering (QA) architectures and techniques.
- Representation learning / Semantic representation models for Visual QA.
- Machine learning methods for Visual QA.
- Zero and one-shot learning methods.
- Use of Web Data and Knowledge Graphs to support Visual QA.
- Linguistic resources and datasets.
- New evaluation paradigms.

Timeline:

* Publication of the training data: December 1st , 2017.
* Challenge papers submission deadline: February 4th, 2018.
* Challenge papers acceptance notification: February 14th, 2018.
* Challenge test data published and submission of results: February 14th, 2018.

Submission:

The Web Conference Challenges is an official track of the conference. We request from participants to provide, in addition to their participation to the challenge, a paper describing the proposed solution and, when relevant, self-assessments related to the defined criteria for evaluation. These papers will be published in the official satellite proceedings of the conference.

Organizers:

Fabricio Faria, Federal University of Rio de Janeiro
Andre Freitas, University of Manchester
Ricardo Usbeck, Paderborn University
Tingting Mu, University of Manchester
Alessio Sarullo, University of Manchester