DeepStyle: Multimodal Search Engine for Fashion and Interior Design

Tautkute, Ivona; Trzciński, Tomasz Piotr; Skorupa, Aleksander; Brocki, Łukasz; Marasek, Krzysztof

doi:10.48550/arxiv.1801.03002

Cited by 3 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To address this limitation, several research groups have worked on incorporating visual information in fashion recommendation systems, mainly with the purpose of recommending similar items to an image query [3,8,14,23,28], and recommending aesthetics based on personal preferences [4,26]. Similarity based fashion recommendation systems are useful for finding substitutes for an item (e.g., finding a shirt with the same style but different brand or price) or matching street images to online products [3,5].…”

Section: Fashion Recommendationmentioning

confidence: 99%

Fashion Recommendation and Compatibility Prediction Using Relational Network

Moosaei¹,

Lin²,

Yang³

2020

Preprint

View full text Add to dashboard Cite

Fashion is an inherently visual concept and computer vision and artificial intelligence (AI) are playing an increasingly important role in shaping the future of this domain. Many research has been done on recommending fashion products based on the learned user preferences. However, in addition to recommending single items, AI can also help users create stylish outfits from items they already have, or purchase additional items that go well with their current wardrobe. Compatibility is the key factor in creating stylish outfits from single items. Previous studies have mostly focused on modeling pair-wise compatibility. There are a few approaches that consider an entire outfit, but these approaches have limitations such as requiring rich semantic information, category labels, and fixed order of items. Thus, they fail to effectively determine compatibility when such information is not available. In this work, we adopt a Relation Network (RN) to develop new compatibility learning models, Fashion RN and FashionRN-VSE, that addresses the limitations of existing approaches. FashionRN learns the compatibility of an entire outfit, with an arbitrary number of items, in an arbitrary order. We evaluated our model using a large dataset of 49,740 outfits that we collected from Polyvore website. Quantitatively, our experimental results demonstrate state of the art performance compared with alternative methods in the literature in both compatibility prediction and fill-in-the-blank test. Qualitatively, we also show that the item embedding learned by FashionRN indicate the compatibility among fashion items. CCS CONCEPTS• Information systems → Collaborative and social computing systems and tools; • Computing methodologies → Artificial intelligence; • Applied computing → Arts and humanities.

show abstract

Section: Fashion Recommendationmentioning

confidence: 99%

Fashion Recommendation and Compatibility Prediction Using Relational Network

Moosaei¹,

Lin²,

Yang³

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Learning visual compatibility is an important problem in fashion recommendation. One possible solution is to learn the compatibility between a pair of fashion items using metric learning [12,14,[18][19][20].…”

Section: Related Work 21 Fashion Compatibility Learning and Out T Gen...mentioning

confidence: 99%

“…To measure the compatibility between items, McAuley et al [12] proposed a method of learning the relation between image features extracted by a pretrained convolutional neural network (CNN). Using a Siamese network, this feature extraction technique for compatibility learning was improved [18,20]. ese methods can learn complex relationships by merely providing samples of positive and negative pairs.…”

Section: Related Work 21 Fashion Compatibility Learning and Out T Gen...mentioning

confidence: 99%

Outfit Generation and Style Extraction via Bidirectional LSTM and Autoencoder

Nakamura¹,

Goto²

2018

Preprint

View full text Add to dashboard Cite

When creating an out t, style is a criterion in selecting each fashion item.is means that style can be regarded as a feature of the overall out t. However, in various previous studies on out t generation, there have been few methods focusing on global information obtained from an out t. To address this de ciency, we have incorporated an unsupervised style extraction module into a model to learn out ts. Using the style information of an out t as a whole, the proposed model succeeded in generating out ts more exibly without requiring additional information . Moreover, the style information extracted by the proposed model is easy to interpret. e proposed model was evaluated on two human-generated out t datasets. In a fashion item prediction task (missing prediction task), the proposed model outperformed a baseline method. In a style extraction task, the proposed model extracted some easily distinguishable styles. In an out t generation task, the proposed model generated an out t while controlling its styles . is capability allows us to generate fashionable out ts according to various preferences. CCS CONCEPTS•Computing methodologies → Image representations; Factorization methods; Neural networks;

show abstract

“…RELATED WORKFashion is an important application domain of computer vision and multimedia. Much research e ort has been made in this domain, focusing on fashion image retrieval[2,32], clothing recognition[16], clothing parsing[12,33], a ribute learning[5,15], out t compatibility[3,17,27,28,30], and fashion recommendation[4,8,15]. e goal of this work is to compose personalized fashion out ts automatically based on user behaviors, we hence focus on the research areas of out t generation and recommendation 4.…”

mentioning

confidence: 99%

POG: Personalized Outfit Generation for Fashion Recommendation at Alibaba iFashion

Chen

Huang

et al. 2019

Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining

139

View full text Add to dashboard Cite

Increasing demand for fashion recommendation raises a lot of challenges for online shopping platforms and fashion communities. In particular, there exist two requirements for fashion out t recommendation: the Compatibility of the generated fashion out ts, and the Personalization in the recommendation process. In this paper, we demonstrate these two requirements can be satis ed via building a bridge between out t generation and recommendation. rough large data analysis, we observe that people have similar tastes in individual items and out ts. erefore, we propose a Personalized Out t Generation (POG) model, which connects user preferences regarding individual items and out ts with Transformer architecture. Extensive o ine and online experiments provide strong quantitative evidence that our method outperforms alternative methods regarding both compatibility and personalization metrics. Furthermore, we deploy POG on a platform named Dida in Alibaba to generate personalized out ts for the users of the online application iFashion.is work represents a rst step towards an industrial-scale fashion out t generation and recommendation solution, which goes beyond generating out ts based on explicit queries, or merely recommending from existing out t pools. As part of this work, we release a large-scale dataset consisting of 1.01 million out ts with rich context information, and 0.28 billion user click actions from 3.57 million users. To the best of our knowledge, this dataset is the largest, publicly available, fashion related dataset, and the rst to provide user behaviors relating to both out ts and fashion items.

show abstract

DeepStyle: Multimodal Search Engine for Fashion and Interior Design

Cited by 3 publications

References 0 publications

Fashion Recommendation and Compatibility Prediction Using Relational Network

Fashion Recommendation and Compatibility Prediction Using Relational Network

Outfit Generation and Style Extraction via Bidirectional LSTM and Autoencoder

POG: Personalized Outfit Generation for Fashion Recommendation at Alibaba iFashion

Contact Info

Product

Resources

About