From Silicon Valley to Lido in Venice: the International Conference on Computer Vision arrives for the first time in Italy
The AI-Computer Vision roadshow continues… From Beijing to Venice. Our fashion discovery research work: Multi-modal embedding for main product detection in fashion, one of the hot topics of ICCV 2017, the top event worldwide reference on Computer Vision.
From October 22nd-29th, the Venue of the Venice Film Palace Congress will host the International Conference on Computer Vision (ICCV 2017). The worldwide reference on Computer Vision, which expects more than 3,000 participants and 60 industrial exhibitors, including Facebook, Apple, Microsoft, Google and Amazon. In addition to the presentation of more than 600 scientific publications and 44 workshops including Computer Vision For Fashion (CVF) organized by the Zalando Research Lab, in which Wide Eyes Technologies will participate.
AI – Computer Vision meets fashion at ICCV 2017 (October 29th, Venice)
The aim of the CVF Workshop powered by Zalando 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 within the fashion industry.
Participants for this workshop will consist of the top experts in AI-Computer Vision in the fashion industries like Prof. Kota Yamaguchi, Research Scientist at CyberAgent, Inc., Dr. Robinson Piramuthu Director of Computer Vision at eBay, and Dr. Edgar Simo-Serra (Research at Waseda University).
The latter will be sharing the principle ideas of fashion discovery research work: Multi-modal embedding for main product detection in fashion with participants. The research work about fashion visual search of Antonio Rubio, Senior Computer Vision Researcher at Wide Eyes Technologies and research scientists at Institut de Robòtica i Informàtica Industrial (CSIC-UPC), was developed in conjunction with Long Long Yu, Co-founder and Head of Research at Wide Eyes Technologies, Edgar Simo-Serra from Waseda University, and Francesc Moreno Noguer, Associate Research at IRI-UPC.
The way the products are presented to the customer is a key factor to increase online sales. In the case of fashion e-commerce, a specific item being sold is normally depicted as worn by a model and tastefully combined with other garments to make it look more attractive. In addition, every product contains textual information or metadata that can be very useful for ceratin tasks.
Following their previous paper on the creation of a multi-modal embedding for fashion product retrieval, presented at this year’s International Conference on Image Processing in Beijing, they apply a similar idea in this new work, while focusing on the main product detection task. They propose to leverage the metadata information to select the most relevant region in an image, or more specifically, to detect the main product in a fashion image that might contain several garments. This allows for the training of specific product classifiers, which do not need to be fed with the whole image.
Last News from ICCV 2017
We are so happy to announce that our fashion discovery research work: Multi-modal embedding for main product detection in fashion won the Best Paper Award of the Computer Vision of Fashion Workshop at ICCV 2017. Special congratulations to our research team.
Good job Antonio Rubio and Long Long Yu!