Paper Track

Accepted Papers

Deep Learning for Automated Tagging of Fashion Images, Patricia Gutierrez, Pierre-Antoine Sondag, Petar Butkovic, Mauro Lacy, Jordi Berges, Felipe Bertrand, and Arne Knudson, [pdf]

Walk Like A Supermodel: Fashion Generation with Pose Adaptation, Sijie Song, Wei Zhang, Jiaying Liu, Tao Mei [pdf]

Everybody Dance Now, Caroline Chan, Shiry Ginosar, Tinghui Zhou, and Alexei A. Efros [pdf]

Brand > Logo: Visual Analysis of Fashion Brands, M. Hadi Kiapour and Robinson Piramuthu, [pdf]

Tiered Deep Similarity Search for Fashion, Dipu Manandhar, Muhammet Bastan, and Kim-Hui Yap, [pdf][supp]

Deep Fashion Analysis with Feature Map Upsampling and Landmark-driven Attention, Jingyuan Liu, Hong Lu, [pdf]

DesIGN: Design Inspiration from Generative Networks, Othman Sbai, Mohamed Elhoseiny, Antoine Bordes, Yann LeCun, Camille Couprie, [pdf] -- Best Paper 🏆

Creating Capsule Wardrobes from Fashion Images, Wei-Lin Hsiao, Kristen Grauman, [pdf]

FashionSearchNet: Fashion Search with Attribute Manipulation, Kenan E. Ak, Ashraf A. Kassim, Joo Hwee Lim, Jo Yew Tham, [pdf]

CRAFT: Complementary Recommendation by Adversarial Feature Transform, Cong Phuoc Huynh, Arridhana Ciptadi, Ambrish Tyagi, Amit Agrawal, [pdf] -- Best Paper Runner-up 🏆

Language Guided Fashion Image Manipulation with Feature-wise Transformations, Mehmet GĂĽnel, Erkut Erdem, Aykut Erdem [pdf]

Full-body High-resolution Anime Generation with Progressive Structure-conditional Generative Adversarial Networks, Koichi Hamada, Kentaro Tachibana, Tianqi Li, Hiroto Honda, and Yusuke Uchida, [pdf] [supp]

Convolutional Photomosaic Generation via Multi-Scale Perceptual Losses, Matthew Tesfaldet , Nariman Saftarli , Marcus A. Brubaker, Konstantinos G. Derpanis, [pdf]

Presentation Guidelines

All accepted papers will be presented as posters. The posters have to be in portrait mode. The poster boards are 1.20x1.00 meters and cannot hold landscape posters.

Submission Guidelines

We solicit short papers on developing and applying computer vision techniques that are valuable in the creative domains, with an emphasis on fashion, art and design. Submitted manuscript should follow the ECCV 2018 paper template. The page limit is less than six pages (excluding references). Submitted work can be extended abstract of recent publications * (including full papers presented at the main conference) or work in progress which may be submitted to a different conference at a later date*. The accepted papers will be linked at the workshop webpage. Papers will be peer reviewed under double-blind policy, and must be submitted online through the CMT submission system at: https://cmt3.research.microsoft.com/CVFAD2018

Important Dates

  • Paper submission deadline: 15 July 2018*
  • Notification to authors: 30 July 2018
  • Camera ready: 30 September 2018

* All accepted papers will be linked at the workshop website. If the paper is not from an already published/accepted work (ECCV, CVPR, etc.), the authors can choose to have their papers included in the ECCV workshop proceedings (published by Springer).

* For authors who want to submit their accepted work at this workshop to a different journal or conference, please check their double submission rules thoroughly. For example, CVPR considers peer-reviewed workshop papers longer than 4 pages as a publication. So if you want to submit your work to CVPR, you need to avoid substantial overlap of content.

*For authors who need to apply European Visa to attend the conference, if you prefer a faster paper review process, please send us an email for this special request.

Topics Covered

Topics of the papers include but are not limited to:

  • Virtual try-on/wardrobe: projects fashion articles onto humans/avatars in order to visualize the realistic fit of an outfit, generative models for 3D.
  • Cross-domain visual search: robust visual search between different domains, such as street-style photos, catalog images and art/design sketches, etc..
  • Personal shopping assistant: multi-modal interaction between the customer and an intelligent agent which aims to assist the shopper’s experience, such as style discovery and accurate product search.
  • Automatic article tagging: automatically tag articles with visual features which enables fast inventory logging.
  • Trend analysis and forecast: automated visual style discovery and trend analysis and forecast from social media data with weak supervision.
  • Efficient fashion search: retrieves images of fashion items that match the online shopper’s query (the query itself can take many forms).
  • Fashion analysis from videos: efficient fashion outfits parsing and retrieval in videos.
  • Style recommendation: suggests fashion articles or outfits which complement the style of a particular article or according to the customer’s preference.
  • Novel methods for visual content generation: including new theoretical insights, new techniques for visual content generation, including images, sketches, videos, 3D content, etc.
  • New evaluation metrics: objective metrics used to compare image generation algorithms on the quality and aesthetic value of generated content.
  • Design with humans in the loop: generative design algorithms which aim to explore systems that can augment human’s creative process, with new approaches to human algorithm interaction.
  • Novel applications: novel applications of image generation that can potentially bring value to the creative domains, such as art, fashion and design.