Fashion Outfit Style Retrieval Based on Hashing Method

This paper proposes an outfit retrieval method for moving fashion items dressed by models in the catwalk videos. The proposed method aims at retrieving similar style in fashion outfits images using a bilinear supervised hashing algorithm. The targeted images are labeled with the information of both color and attributes for comprehensive description of clothing style. The speed up robust features (SURFs) are extracted as the low-level features of fashion images and fed into the hashing algorithm as original data. To achieve better retrieval performance, a bilinear supervised hashing method is employed to learn high-quality hash codes. Outfits with similar style to the target are expected to be retrieved by hash code ranking. The experiment was conducted on a dataset composed of fashion outfit images and the experimental results show that the proposed method can retrieve styles which are close to query of color and attributes.

[1]  Liang Lin,et al.  Clothing Co-parsing by Joint Image Segmentation and Labeling , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[2]  Changsheng Xu,et al.  Street-to-shop: Cross-scenario clothing retrieval via parts alignment and auxiliary set , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Lianhong Cai,et al.  Interpretable aesthetic features for affective image classification , 2013, 2013 IEEE International Conference on Image Processing.

[4]  Yejun Liu,et al.  Towards Better Understanding the Clothing Fashion Styles: A Multimodal Deep Learning Approach , 2017, AAAI.

[5]  Robinson Piramuthu,et al.  Style Finder: Fine-Grained Clothing Style Detection and Retrieval , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops.

[6]  Jie Xu,et al.  A web-based design support system for fashion technical sketches , 2016 .

[7]  Sanjiv Kumar,et al.  Learning Binary Codes for High-Dimensional Data Using Bilinear Projections , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[8]  Noah Snavely,et al.  StreetStyle: Exploring world-wide clothing styles from millions of photos , 2017, ArXiv.

[9]  Wai Keung Wong,et al.  Supervised discrete discriminant hashing for image retrieval , 2018, Pattern Recognit..

[10]  Zhiyuan Liu,et al.  Learning to Appreciate the Aesthetic Effects of Clothing , 2016, AAAI.

[11]  Hanqing Lu,et al.  Street-to-shop: Cross-scenario clothing retrieval via parts alignment and auxiliary set , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[12]  Bruno Lévy,et al.  Geometry-aware direction field processing , 2009, TOGS.

[13]  Wei Liu,et al.  Supervised Discrete Hashing , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Nicu Sebe,et al.  A Survey on Learning to Hash , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  Alexander C. Berg,et al.  Hipster Wars: Discovering Elements of Fashion Styles , 2014, ECCV.

[16]  Hiroshi Ishikawa,et al.  Fashion Style in 128 Floats: Joint Ranking and Classification Using Weak Data for Feature Extraction , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Huizhong Chen,et al.  Describing Clothing by Semantic Attributes , 2012, ECCV.

[18]  Jian Dong,et al.  Deep domain adaptation for describing people based on fine-grained clothing attributes , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  Antonio Torralba,et al.  Spectral Hashing , 2008, NIPS.

[20]  Xiaogang Wang,et al.  DeepFashion: Powering Robust Clothes Recognition and Retrieval with Rich Annotations , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[21]  Chun Chen,et al.  Data-driven image color theme enhancement , 2010, SIGGRAPH 2010.