skip to main content
Ngôn ngữ:
Giới hạn tìm kiếm: Giới hạn tìm kiếm: Dạng tài nguyên Hiển thị kết quả với: Hiển thị kết quả với: Chỉ mục

Adaptive Kernel Size Selection for Correntropy Based Metric

Tan, Ying ; Fang, Yuchun ; Li, Yang ; Dai, Wang; Park, Jong-Il (Editor) ; Kim, Junmo (Editor)

Lecture Notes in Computer Science, Computer Vision - ACCV 2012 Workshops: ACCV 2012 International Workshops, Daejeon, Korea, November 5-6, 2012, Revised Selected Papers, Part I, pp.50-60

ISBN: 9783642374098 ; ISBN: 3642374093 ; E-ISBN: 9783642374104 ; E-ISBN: 3642374107 ; DOI: 10.1007/978-3-642-37410-4_5

Toàn văn sẵn có

Trích dẫn Trích dẫn bởi
  • Nhan đề:
    Adaptive Kernel Size Selection for Correntropy Based Metric
  • Tác giả: Tan, Ying ; Fang, Yuchun ; Li, Yang ; Dai, Wang
  • Park, Jong-Il (Editor) ; Kim, Junmo (Editor)
  • Chủ đề: Computer Science ; Computer Imaging, Vision, Pattern Recognition and Graphics ; User Interfaces and Human Computer Interaction ; Media Design ; Geometry ; Engineering ; Applied Sciences ; Computer Science
  • Là 1 phần của: Lecture Notes in Computer Science, Computer Vision - ACCV 2012 Workshops: ACCV 2012 International Workshops, Daejeon, Korea, November 5-6, 2012, Revised Selected Papers, Part I, pp.50-60
  • Mô tả: The correntropy is originally proposed to measure the similarity between two random variables and developed as a novel metrics for feature matching. As a kernel method, the parameter of kernel function is very important for correntropy metrics. In this paper, we propose an adaptive parameter selection strategy for correntropy metrics and deduce a close-form solution based on the Maximum Correntropy Criterion (MCC). Moreover, considering the correlation of localized features, we modify the classic correntropy into a block-wise metrics. We verify the proposed metrics in face recognition applications taking Local Binary Pattern (LBP) features. Combined with the proposed adaptive parameter selection strategy, the modified block-wise correntropy metrics could result in much better performance in the experiments.
  • Nơi xuất bản: Berlin, Heidelberg: Springer Berlin Heidelberg
  • Năm xuất bản: 2013
  • Ngôn ngữ: English
  • Số nhận dạng: ISBN: 9783642374098 ; ISBN: 3642374093 ; E-ISBN: 9783642374104 ; E-ISBN: 3642374107 ; DOI: 10.1007/978-3-642-37410-4_5

Đang tìm Cơ sở dữ liệu bên ngoài...