Registered Data

[CT197]


  • Session Time & Room
    • CT197 (1/1) : 4E @A502 [Chair: Zhigang Jia]
  • Classification
    • CT197 (1/1) : Communication, information (94A)

[02590] Non-Local Robust Quaternion Matrix Completion for Large-Scale Color Images and Videos Inpainting

  • Session Time & Room : 4E (Aug.24, 17:40-19:20) @A502
  • Type : Contributed Talk
  • Abstract : The image nonlocal self-similarity (NSS) prior refers to the fact that a local patch often has many nonlocal similar patches to it across the image. In this talk we apply such NSS prior to enhance the robust quaternion matrix completion (QMC) method and significantly improve the inpainting performance. A patch group based NSS prior learning scheme is proposed to learn explicit NSS models from natural color images. The NSS-based QMC algorithm computes an optimal low-rank approximation to the high-rank color image, resulting in high PSNR and SSIM measures and particularly the better visual quality. A new joint NSS-base QMC method is also presented to solve the color video inpainting problem based quaternion tensor representation. The numerical experiments on large-scale color images and videos indicate the advantages of NSS-based QMC over the state-of-the-art methods.
  • Classification : 94A08, 68U10
  • Format : Talk at Waseda University
  • Author(s) :
    • Zhigang Jia (Jiangsu Normal University)

[02595] Image recovery under non-Gaussian noise

  • Session Time & Room : 4E (Aug.24, 17:40-19:20) @A502
  • Type : Contributed Talk
  • Abstract : Cauchy noise, as a typical non-Gaussian noise, appears frequently in many important fields, such as radar, medical, and biomedical imaging. Here, we focus on image recovery under Cauchy noise. Instead of the celebrated total variation or low-rank prior, we adopt a novel deep-learning-based image denoiser prior to effectively remove Cauchy noise with blur. To preserve more detailed texture and better balance between the receptive field size and the computational cost, we apply the multi-level wavelet convolutional neural network (MWCNN) to train this denoiser. Frequently appearing in medical imaging, Rician noise leads to an interesting nonconvex optimization problem, termed as the MAP-Rician model, which is based on the Maximum a Posteriori (MAP) estimation approach. As the MAP-Rician model is deeply rooted in Bayesian analysis, we want to understand its mathematical analysis carefully. Moreover, one needs to properly select a suitable algorithm for tackling this nonconvex problem to get the best performance. Indeed, we first present a theoretical result about the existence of a minimizer for the MAP-Rician model under mild conditions. Next, we aim to adopt an efficient boosted difference of convex functions algorithm (BDCA) to handle this challenging problem. Theoretically, using the Kurdyka-Lojasiewicz (KL) property, the convergence of the numerical algorithm can be guaranteed.
  • Classification : 94A08, 68U10
  • Format : Talk at Waseda University
  • Author(s) :
    • Tingting WU (Nanjing University of Posts and Telecommunications)

[00578] Secret Sharing Scheme with Perfect Concealment by Quasigroup

  • Session Time & Room : 4E (Aug.24, 17:40-19:20) @A502
  • Type : Contributed Talk
  • Abstract : A secret sharing scheme was introduced by Shamir in 1979. A quasigroup is equivalent to a Latin square. The concept of perfect concealment is also called perfect security. The word ‘security’ describes a property of a phenomena, and the word ‘concealment’ describes an action which makes a phenomenon. In this talk, we force an action rather than a property, and we give new construction of secret sharing scheme with perfect concealment by quasigroup.
  • Classification : 94A62, 05B15, 20N05, 60B99
  • Format : Talk at Waseda University
  • Author(s) :
    • Tomoko Adachi (Shizuoka Institute of Science and Technology)
    • Izumi Takeuti (National Institute of Advanced Industrial Science and Technology)

[02380] PDE methods for joint reconstruction-segmentation of images

  • Session Time & Room : 4E (Aug.24, 17:40-19:20) @A502
  • Type : Contributed Talk
  • Abstract : In practical image segmentation tasks, the image must first be reconstructed from indirect/damaged/noisy observations. Traditionally, reconstruction-segmentation would be performed in sequence: first reconstruct, then segment. Joint reconstruction-segmentation performs reconstruction and segmentation simultaneously, using each to guide the other. Past joint reconstruction-segmentation has employed relatively simple segmentation algorithms, e.g. Chan–Vese. This talk will describe how joint reconstruction-segmentation can be performed using Bhattacharyya-flow-based segmentation (Michailovich et al., 2007) and graph-PDE-based segmentation (Merkurjev et al., 2013).
  • Classification : 94A08, 35Q93, 35R02
  • Format : Talk at Waseda University
  • Author(s) :
    • Jeremy Michael Budd (California Institute of Technology )
    • Franca Hoffmann (California Institute of Technology )
    • Allen Tannenbaum (Stony Brook University)
    • Yves van Gennip (Technische Universiteit Delft)
    • Carola-Bibiane Schönlieb (University of Cambridge)
    • Jonas Latz (Heriot-Watt University)

[02139] Normalizing Flows Based Mutual Information Estimation

  • Session Time & Room : 4E (Aug.24, 17:40-19:20) @A502
  • Type : Contributed Talk
  • Abstract : Mutual Information is a measure of mutual dependence on random quantities without specific modelling assumptions. However, estimating mutual information numerically from high-dimensional data remains a difficult problem. We propose a principled mutual information estimator based on a generalization of normalizing flows. The proposed method uses an autoregressive structure in estimating mutual information with estimating marginal and joint entropy simultaneously. Empirical results demonstrate that our proposed estimator exhibits improved bias-variance trade-offs on standard benchmark tasks.
  • Classification : 94a17, 62b10, 68t07, 68t09
  • Format : Talk at Waseda University
  • Author(s) :
    • Haoran Ni (University of Warwick)
    • Martin Lotz (University of Warwick)