List of Articles Variational Bayes

• Open Access Article
  • Abstract Page
  • Full-Text
  
  1 - Wavelet-based Bayesian Algorithm for Distributed Compressed Sensing
  Razieh Torkamani Ramezan Ali Sadeghzadeh
  10.7508/jist.2019.02.001
  The emerging field of compressive sensing (CS) enables the reconstruction of the signal from a small set of linear projections. Traditional CS deals with a single signal; while one can jointly reconstruct multiple signals via distributed CS (DCS) algorithm. DCS inversio More

  The emerging field of compressive sensing (CS) enables the reconstruction of the signal from a small set of linear projections. Traditional CS deals with a single signal; while one can jointly reconstruct multiple signals via distributed CS (DCS) algorithm. DCS inversion method exploits both the inter- and intra-signal correlations via joint sparsity models (JSM). Since the wavelet coefficients of many signals is sparse, in this paper, the wavelet transform is used as sparsifying transform, and a new wavelet-based Bayesian DCS algorithm (WB-DCS) is proposed, which takes into account the inter-scale dependencies among the wavelet coefficients via hidden Markov tree model (HMT), as well as the inter-signal correlations. This paper uses the Bayesian procedure to statistically model this correlations via the prior distributions. Also, in this work, a type-1 JSM (JSM-1) signal model is used for jointly sparse signals, in which every sparse coefficient vector is considered as the sum of a common component and an innovation component. In order to jointly reconstruct multiple sparse signals, the centralized approach is used in DCS, in which all the data is processed in the fusion center (FC). Also, variational Bayes (VB) procedure is used to infer the posterior distributions of unknown variables. Simulation results demonstrate that the structure exploited within the wavelet coefficients provides superior performance in terms of average reconstruction error and structural similarity index. Manuscript profile
• Open Access Article
  • Abstract Page
  • Full-Text
  
  2 - Variational Bayesian inference in Noise Removal from Hyperspectral Images Using Cluster-Based Latent Variables
  T. Bahraini Abass Ebrahimi moghadam M. Khademi H. Sadoghi Yazdi
  Removing noise from hyperspectral images is an inevitable step to improve the quality of these types of images. Many methods have been proposed by researchers in this field. Most of these methods do not address simultaneous spatial-spectral similarities. When the noise More

  Removing noise from hyperspectral images is an inevitable step to improve the quality of these types of images. Many methods have been proposed by researchers in this field. Most of these methods do not address simultaneous spatial-spectral similarities. When the noise removal method applies data globally without regard to spatial-spectral similarities, it usually has a negative effect on low-level pixels; when in the spectral data, a large number of pixels have little noise and a small number of pixels are destroyed by the high level of noise. In this paper, we first extract spatial-spectral similarities in images by defining cluster-based latent variables. In the following, a low-rank matrix factorization method based on these latent variables is proposed to eliminate the noise of hyperspectral images and to improve the resistance to noise (as compared to other methods). The performance of the proposed method is compared visually with six new methods on real noise-contaminated images. For quantitative comparison, the same experiments are done on clean images combined with six types of simulated noise. The simulation results show that by applying latent variables in the Bayesian inference framework, the performance of the noise removal method is improved and the proposed method performs better than the other methods. Manuscript profile