Registered Data

[CT125]

[01464] Fractional norm regularization under a generative deep learning framework for image restoration

  • Session Date & Time : 3C (Aug.23, 13:20-15:00)
  • Type : Contributed Talk
  • Abstract : Fractional norm regularization has been extensively used in various scientific applications. The usual integer norms are not sufficient to retains many important information inherent in the underlying images. In this study, we intent to incorporate fractional norms in the loss function of a generative model designed to handle various distortion present in the input images. A variational framework using a fractional norm has to be designed to handle various kinds of distortions in the data.
  • Classification : 68U10, 62H35, 15A29, 35A15, 47A52
  • Author(s) :
    • Jidesh Pacheeri Padikkal (National Institute of Technology Karnataka, Surathkal)
    • Bini A A (National Institute of Technology Calicut)

[02201] Total Variation and Undecimated Wavelet Approach to Chest Radiograph Image Enhancement

  • Session Date & Time : 3C (Aug.23, 13:20-15:00)
  • Type : Contributed Talk
  • Abstract : ABSTRACT Medical images are often corrupted by white noise, blurring and contrast defects. Consequently, important medical information may be degraded or completely masked. In this talk, I will discuss recent advances in the implementation of a total variation method and an undecimated wavelet image enhancing algorithm, as well as morphological techniques such as closing, thinning and pruning combined with wavelets methods, to improve the image quality and to extract target features from chest radiographs.
  • Classification : 68U10, 44A20, 46N10, Chest radiograph image enhancement, mathematical morphology, wavelet decomposition
  • Author(s) :
    • Anthony Y Aidoo (Eastern Connecticut State University)
    • Gloria Agyeiwaa Botchway (University of Ghana)
    • Matilda Acheampong Wilson (University of Ghana)

[02586] Quaternion-valued Wiener Filters: Applications to color imaging denoising

  • Session Date & Time : 3C (Aug.23, 13:20-15:00)
  • Type : Contributed Talk
  • Abstract : This research investigates the use of quaternion-valued Wiener filters (QWF) for color image denoising, leveraging the unique properties of the quaternion number system to consider inter-channel correlations. The proposed method is applied to a range of noisy color images, and performance is compared with traditional denoising techniques. Results indicate that QWF provide superior performance in terms of noise reduction and preservation of image details, demonstrating the potential of this approach for advanced color imaging applications.
  • Classification : 68U10, 94A08, 94A12, 15A69
  • Author(s) :
    • ANTONIO MARQUINA (UNIVERSITY OF VALENCIA)
    • Víctor Toledo Prades (University of Valencia)

[00077] Machine Learning challenges to automate behavioral scales

  • Session Date & Time : 3C (Aug.23, 13:20-15:00)
  • Type : Industrial Contributed Talk
  • Abstract : Psychologists often use short speech samples to classify patients according to various behavioral scales, such as depression, suicidality, and mania. We are investigating the ability of Machine Learning techniques to replicate the scores of the psychologist produced scores with the goal of tracking longitudinal behavior without the need of frequent in-person interactions. In this talk we present our results and challenges.
  • Classification : 68U15, 91E45
  • Author(s) :
    • Filippo Posta (Estrella Mountain CC)

[02185] Vibration of anti-symmetric angle-ply layered conical shell frusta using splines

  • Session Date & Time : 3C (Aug.23, 13:20-15:00)
  • Type : Industrial Contributed Talk
  • Abstract : Vibration of anti-symmetric angle-ply layered conical shell is studied including shear deformation using spline function approximation. The equilibrium equations are formulated in terms of displacement and rotational functions. These functions are approximated using Bickley-type splines to obtain the generalised eigenvalue problem. Parametric studies are made to analyse the effects of circumferential node number, length ratio and cone angle on the frequency parameter for number of layers and materials with ply orientations under different boundary conditions.
  • Classification : 74K25, 74-10, 74E30, 74G15, 74B05, Composite Shell structure
  • Author(s) :
    • Viswanathan Kodakkal Kannan (Samarkand State University)