Texture retrieval using mixtures of generalized Gaussian distribution and Cauchy–Schwarz divergence in wavelet domain
作者：
Highlights：
• We propose Cauchy Schwarz divergence CSD for texture retrieval.
• Wavelet coefficients are modeled by mixture of generalized Gaussians MoGG.
• We derive a closed-form of CSD between MoGG for fixed parameter shape.
• We use the Monte-Carlo approximation for CSD in the general case.
• We evaluate the performance in terms of the average retrieval rates and the computational time.
摘要
Highlights•We propose Cauchy Schwarz divergence CSD for texture retrieval.•Wavelet coefficients are modeled by mixture of generalized Gaussians MoGG.•We derive a closed-form of CSD between MoGG for fixed parameter shape.•We use the Monte-Carlo approximation for CSD in the general case.•We evaluate the performance in terms of the average retrieval rates and the computational time.
论文关键词：Wavelet decomposition,Mixture of generalized Gaussian model,Similarity measurement,Cauchy–Schwarz divergence
论文评审过程：Received 29 January 2015, Revised 6 January 2016, Accepted 12 January 2016, Available online 21 January 2016, Version of Record 5 February 2016.
论文官网地址：https://doi.org/10.1016/j.image.2016.01.005
