Interpretable partitioned embedding for customized multi-item fashion outfit composition

Zunlei Feng, Zhenyun Yu, Yezhou Yang, Yongcheng Jing, Junxiao Jiang, Mingli Song

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Intelligent fashion outfit composition becomes more and more popular in these years. Some deep learning based approaches reveal competitive composition recently. However, the uninterpretable characteristic makes such deep learning based approach cannot meet the designers, businesses and consumers' urge to comprehend the importance of different attributes in an outfit composition. To realize interpretable and customized multi-item fashion outfit compositions, we propose a partitioned embedding network to learn interpretable embeddings from clothing items. The network consists of two vital components: attribute partition module and partition adversarial module. In the attribute partition module, multiple attribute labels are adopted to ensure that different parts of the overall embedding correspond to different attributes. In the partition adversarial module, adversarial operations are adopted to achieve the independence of different parts. With the interpretable and partitioned embedding, we then construct an outfit composition graph and an attribute matching map. Extensive experiments demonstrate that 1) the partitioned embedding have unmingled parts which corresponding to different attributes and 2) outfits recommended by our model are more desirable in comparison with the existing methods.

Original languageEnglish (US)
Title of host publicationICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval
PublisherAssociation for Computing Machinery, Inc
Pages143-151
Number of pages9
ISBN (Print)9781450350464
DOIs
StatePublished - Jun 5 2018
Event8th ACM International Conference on Multimedia Retrieval, ICMR 2018 - Yokohama, Japan
Duration: Jun 11 2018Jun 14 2018

Other

Other8th ACM International Conference on Multimedia Retrieval, ICMR 2018
CountryJapan
CityYokohama
Period6/11/186/14/18

Fingerprint

Chemical analysis
Labels
Industry
Experiments
Deep learning

Keywords

  • Adversarial
  • Embedding
  • Interpretable
  • Outfit composition

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Networks and Communications
  • Computer Science Applications

Cite this

Feng, Z., Yu, Z., Yang, Y., Jing, Y., Jiang, J., & Song, M. (2018). Interpretable partitioned embedding for customized multi-item fashion outfit composition. In ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval (pp. 143-151). Association for Computing Machinery, Inc. https://doi.org/10.1145/3206025.3206048

Interpretable partitioned embedding for customized multi-item fashion outfit composition. / Feng, Zunlei; Yu, Zhenyun; Yang, Yezhou; Jing, Yongcheng; Jiang, Junxiao; Song, Mingli.

ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval. Association for Computing Machinery, Inc, 2018. p. 143-151.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Feng, Z, Yu, Z, Yang, Y, Jing, Y, Jiang, J & Song, M 2018, Interpretable partitioned embedding for customized multi-item fashion outfit composition. in ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval. Association for Computing Machinery, Inc, pp. 143-151, 8th ACM International Conference on Multimedia Retrieval, ICMR 2018, Yokohama, Japan, 6/11/18. https://doi.org/10.1145/3206025.3206048
Feng Z, Yu Z, Yang Y, Jing Y, Jiang J, Song M. Interpretable partitioned embedding for customized multi-item fashion outfit composition. In ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval. Association for Computing Machinery, Inc. 2018. p. 143-151 https://doi.org/10.1145/3206025.3206048
Feng, Zunlei ; Yu, Zhenyun ; Yang, Yezhou ; Jing, Yongcheng ; Jiang, Junxiao ; Song, Mingli. / Interpretable partitioned embedding for customized multi-item fashion outfit composition. ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval. Association for Computing Machinery, Inc, 2018. pp. 143-151
@inproceedings{99e1406153f34873b7ecf28346612f9c,
title = "Interpretable partitioned embedding for customized multi-item fashion outfit composition",
abstract = "Intelligent fashion outfit composition becomes more and more popular in these years. Some deep learning based approaches reveal competitive composition recently. However, the uninterpretable characteristic makes such deep learning based approach cannot meet the designers, businesses and consumers' urge to comprehend the importance of different attributes in an outfit composition. To realize interpretable and customized multi-item fashion outfit compositions, we propose a partitioned embedding network to learn interpretable embeddings from clothing items. The network consists of two vital components: attribute partition module and partition adversarial module. In the attribute partition module, multiple attribute labels are adopted to ensure that different parts of the overall embedding correspond to different attributes. In the partition adversarial module, adversarial operations are adopted to achieve the independence of different parts. With the interpretable and partitioned embedding, we then construct an outfit composition graph and an attribute matching map. Extensive experiments demonstrate that 1) the partitioned embedding have unmingled parts which corresponding to different attributes and 2) outfits recommended by our model are more desirable in comparison with the existing methods.",
keywords = "Adversarial, Embedding, Interpretable, Outfit composition",
author = "Zunlei Feng and Zhenyun Yu and Yezhou Yang and Yongcheng Jing and Junxiao Jiang and Mingli Song",
year = "2018",
month = "6",
day = "5",
doi = "10.1145/3206025.3206048",
language = "English (US)",
isbn = "9781450350464",
pages = "143--151",
booktitle = "ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval",
publisher = "Association for Computing Machinery, Inc",

}

TY - GEN

T1 - Interpretable partitioned embedding for customized multi-item fashion outfit composition

AU - Feng, Zunlei

AU - Yu, Zhenyun

AU - Yang, Yezhou

AU - Jing, Yongcheng

AU - Jiang, Junxiao

AU - Song, Mingli

PY - 2018/6/5

Y1 - 2018/6/5

N2 - Intelligent fashion outfit composition becomes more and more popular in these years. Some deep learning based approaches reveal competitive composition recently. However, the uninterpretable characteristic makes such deep learning based approach cannot meet the designers, businesses and consumers' urge to comprehend the importance of different attributes in an outfit composition. To realize interpretable and customized multi-item fashion outfit compositions, we propose a partitioned embedding network to learn interpretable embeddings from clothing items. The network consists of two vital components: attribute partition module and partition adversarial module. In the attribute partition module, multiple attribute labels are adopted to ensure that different parts of the overall embedding correspond to different attributes. In the partition adversarial module, adversarial operations are adopted to achieve the independence of different parts. With the interpretable and partitioned embedding, we then construct an outfit composition graph and an attribute matching map. Extensive experiments demonstrate that 1) the partitioned embedding have unmingled parts which corresponding to different attributes and 2) outfits recommended by our model are more desirable in comparison with the existing methods.

AB - Intelligent fashion outfit composition becomes more and more popular in these years. Some deep learning based approaches reveal competitive composition recently. However, the uninterpretable characteristic makes such deep learning based approach cannot meet the designers, businesses and consumers' urge to comprehend the importance of different attributes in an outfit composition. To realize interpretable and customized multi-item fashion outfit compositions, we propose a partitioned embedding network to learn interpretable embeddings from clothing items. The network consists of two vital components: attribute partition module and partition adversarial module. In the attribute partition module, multiple attribute labels are adopted to ensure that different parts of the overall embedding correspond to different attributes. In the partition adversarial module, adversarial operations are adopted to achieve the independence of different parts. With the interpretable and partitioned embedding, we then construct an outfit composition graph and an attribute matching map. Extensive experiments demonstrate that 1) the partitioned embedding have unmingled parts which corresponding to different attributes and 2) outfits recommended by our model are more desirable in comparison with the existing methods.

KW - Adversarial

KW - Embedding

KW - Interpretable

KW - Outfit composition

UR - http://www.scopus.com/inward/record.url?scp=85053923132&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053923132&partnerID=8YFLogxK

U2 - 10.1145/3206025.3206048

DO - 10.1145/3206025.3206048

M3 - Conference contribution

AN - SCOPUS:85053923132

SN - 9781450350464

SP - 143

EP - 151

BT - ICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval

PB - Association for Computing Machinery, Inc

ER -