Computer Vision For Fashion

ICCV 2017 Workshop

Fashion, in its various forms, exercises a large influence on many aspects of modern societies, having a strong financial and cultural impact. This fact in conjunction with the rapid evolution in the field of computer vision has led to an increased interest in the automatic analysis of the strong visual components of fashion. This steady growth in interest has made CV for fashion an emerging and exciting field in its own right. A multitude of fashion related visual systems have been proposed in recent years, each aiming to exploit state-of-the-art algorithms in the fields of Computer Vision and Machine Learning and relevant emerging technologies.

At the same time, these initial attempts have highlighted the peculiarities and particular challenges of the field. Such problems range from the purely technical, e.g. the need for high resolution analysis, to the artistic, e.g recent efforts to create fashion (texture or even articles). A further characteristic of the field is the large diversity of possible applications each with its own challenges; applications can vary from classical item (visual) recommendations, to style recognition, to applications encompassing various components such as visual style assistants.

The aim of the CV for Fashion workshop is to bring together researchers working on various aspects of fashion-related visual analysis and generation, in order to discuss the current state of the field and the various challenges moving forward. There will be a focus on the presentation of both established and novel ideas. Equal importance will be given to issues which many emerging fields face as for instance precise problem definition, evaluation protocols, and benchmarking datasets.

List of Possible Topics

  • Street-to-shop.
  • Article recommendations.
  • Style / outfit recommendations.
  • Visual search for fashion.
  • Virtual personal fashion assistants.
  • Virtual wardrobe.
  • Recommendations with humans in the loop.
  • Automatic article tagging.
  • Retrieval algorithms for large fashion catalogues.
  • Algorithmic clothing.
  • Predicting fashionability scores.

This is not an exhaustive list. We welcome other interesting and relevant research.

Important Dates

Submission deadline: July 7th, 2017

Author notification: August 10th, 2017

Camera-ready due: August 18th, 2017

Workshop date: October 29th, 2017