Deep metric learning with online hard and soft selection for person re-identification

Mingyang Yu, Sei Ichiro Kamata

研究成果: Conference contribution

抄録

Deep metric learning has been widely used for image retrieval and verification tasks. Traditional contrastive loss and triplet loss depend highly on the selection of pair/triplet images. It makes the training process unstable and uncomplete. In this paper, we propose a novel global level loss function that considers histograms for intra distances within class and inter distances between different classes. We compared two forms of global level loss (hard selection based loss and soft selection based loss) and both achieved better result than traditional triplet loss, multi-class N pair loss and other related works. The experiment is conducted on the person re-identification dataset Market 1501 and DukeMTMC-reID.

本文言語English
ホスト出版物のタイトル2018 Joint 7th International Conference on Informatics, Electronics and Vision and 2nd International Conference on Imaging, Vision and Pattern Recognition, ICIEV-IVPR 2018
出版社Institute of Electrical and Electronics Engineers Inc.
ページ426-431
ページ数6
ISBN(電子版)9781538651612
DOI
出版ステータスPublished - 2019 2 12
イベントJoint 7th International Conference on Informatics, Electronics and Vision and 2nd International Conference on Imaging, Vision and Pattern Recognition, ICIEV-IVPR 2018 - Kitakyushu, Japan
継続期間: 2018 6 252018 6 28

出版物シリーズ

名前2018 Joint 7th International Conference on Informatics, Electronics and Vision and 2nd International Conference on Imaging, Vision and Pattern Recognition, ICIEV-IVPR 2018

Conference

ConferenceJoint 7th International Conference on Informatics, Electronics and Vision and 2nd International Conference on Imaging, Vision and Pattern Recognition, ICIEV-IVPR 2018
国/地域Japan
CityKitakyushu
Period18/6/2518/6/28

ASJC Scopus subject areas

  • 信号処理
  • 制御と最適化
  • 人工知能
  • コンピュータ ビジョンおよびパターン認識
  • 情報システム

フィンガープリント

「Deep metric learning with online hard and soft selection for person re-identification」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル