-
Bayesian optimization for sparse neural networks with trainable activation functions
Authors:
Mohamed Fakhfakh,
Lotfi Chaari
Abstract:
In the literature on deep neural networks, there is considerable interest in develo** activation functions that can enhance neural network performance. In recent years, there has been renewed scientific interest in proposing activation functions that can be trained throughout the learning process, as they appear to improve network performance, especially by reducing overfitting. In this paper, w…
▽ More
In the literature on deep neural networks, there is considerable interest in develo** activation functions that can enhance neural network performance. In recent years, there has been renewed scientific interest in proposing activation functions that can be trained throughout the learning process, as they appear to improve network performance, especially by reducing overfitting. In this paper, we propose a trainable activation function whose parameters need to be estimated. A fully Bayesian model is developed to automatically estimate from the learning data both the model weights and activation function parameters. An MCMC-based optimization scheme is developed to build the inference. The proposed method aims to solve the aforementioned problems and improve convergence time by using an efficient sampling scheme that guarantees convergence to the global maximum. The proposed scheme is tested on three datasets with three different CNNs. Promising results demonstrate the usefulness of our proposed approach in improving model accuracy due to the proposed activation function and Bayesian estimation of the parameters.
△ Less
Submitted 19 April, 2023; v1 submitted 10 April, 2023;
originally announced April 2023.
-
Spatio-temporal wavelet regularization for parallel MRI reconstruction: application to functional MRI
Authors:
Lotfi Chaari,
Sébastien Mériaux,
Jean-Christophe Pesquet,
Philippe Ciuciu
Abstract:
Parallel MRI is a fast imaging technique that enables the acquisition of highly resolved images in space or/and in time. The performance of parallel imaging strongly depends on the reconstruction algorithm, which can proceed either in the original k-space (GRAPPA, SMASH) or in the image domain (SENSE-like methods). To improve the performance of the widely used SENSE algorithm, 2D- or slice-specifi…
▽ More
Parallel MRI is a fast imaging technique that enables the acquisition of highly resolved images in space or/and in time. The performance of parallel imaging strongly depends on the reconstruction algorithm, which can proceed either in the original k-space (GRAPPA, SMASH) or in the image domain (SENSE-like methods). To improve the performance of the widely used SENSE algorithm, 2D- or slice-specific regularization in the wavelet domain has been deeply investigated. In this paper, we extend this approach using 3D-wavelet representations in order to handle all slices together and address reconstruction artifacts which propagate across adjacent slices. The gain induced by such extension (3D-Unconstrained Wavelet Regularized -SENSE: 3D-UWR-SENSE) is validated on anatomical image reconstruction where no temporal acquisition is considered. Another important extension accounts for temporal correlations that exist between successive scans in functional MRI (fMRI). In addition to the case of 2D+t acquisition schemes addressed by some other methods like kt-FOCUSS, our approach allows us to deal with 3D+t acquisition schemes which are widely used in neuroimaging. The resulting 3D-UWR-SENSE and 4D-UWR-SENSE reconstruction schemes are fully unsupervised in the sense that all regularization parameters are estimated in the maximum likelihood sense on a reference scan. The gain induced by such extensions is illustrated on both anatomical and functional image reconstruction, and also measured in terms of statistical sensitivity for the 4D-UWR-SENSE approach during a fast event-related fMRI protocol. Our 4D-UWR-SENSE algorithm outperforms the SENSE reconstruction at the subject and group levels (15 subjects) for different contrasts of interest (eg, motor or computation tasks) and using different parallel acceleration factors (R=2 and R=4) on 2x2x3mm3 EPI images.
△ Less
Submitted 3 October, 2013; v1 submitted 23 December, 2011;
originally announced January 2012.
-
4D Wavelet-Based Regularization for Parallel MRI Reconstruction: Impact on Subject and Group-Levels Statistical Sensitivity in fMRI
Authors:
Lotfi Chaari,
Sébastien Mériaux,
Solveig Badillo,
Jean-Christophe Pesquet,
Philippe Ciuciu
Abstract:
Parallel MRI is a fast imaging technique that enables the acquisition of highly resolved images in space. It relies on $k$-space undersampling and multiple receiver coils with complementary sensitivity profiles in order to reconstruct a full Field-Of-View (FOV) image. The performance of parallel imaging mainly depends on the reconstruction algorithm, which can proceed either in the original $k$-sp…
▽ More
Parallel MRI is a fast imaging technique that enables the acquisition of highly resolved images in space. It relies on $k$-space undersampling and multiple receiver coils with complementary sensitivity profiles in order to reconstruct a full Field-Of-View (FOV) image. The performance of parallel imaging mainly depends on the reconstruction algorithm, which can proceed either in the original $k$-space (GRAPPA, SMASH) or in the image domain (SENSE-like methods). To improve the performance of the widely used SENSE algorithm, 2D- or slice-specific regularization in the wavelet domain has been efficiently investigated. In this paper, we extend this approach using 3D-wavelet representations in order to handle all slices together and address reconstruction artifacts which propagate across adjacent slices. The extension also accounts for temporal correlations that exist between successive scans in functional MRI (fMRI). The proposed 4D reconstruction scheme is fully \emph{unsupervised} in the sense that all regularization parameters are estimated in the maximum likelihood sense on a reference scan. The gain induced by such extensions is first illustrated on EPI image reconstruction but also measured in terms of statistical sensitivity during a fast event-related fMRI protocol. The proposed 4D-UWR-SENSE algorithm outperforms the SENSE reconstruction at the subject and group-levels (15 subjects) for different contrasts of interest and using different parallel acceleration factors on $2\times2\times3$mm$^3$ EPI images.
△ Less
Submitted 17 March, 2011;
originally announced March 2011.
-
Conception and FPGA implementation of IEEE 802.11s mesh network MAC layer transmitter
Authors:
Lamia Chaari,
Rim Ayadi,
Lotfi Kamoun
Abstract:
This paper proposes, a hardware implementation of Wireless Mesh Networks (WMN) medium Access Controller (MAC) layer transmitter. In the literature a lot of works are focused on WMN routing protocol as well as performance analysis and software integration of WMN units, however few works deals with WMN hardware implementation. In this field our contribution is to conceive and to implements on FPGA a…
▽ More
This paper proposes, a hardware implementation of Wireless Mesh Networks (WMN) medium Access Controller (MAC) layer transmitter. In the literature a lot of works are focused on WMN routing protocol as well as performance analysis and software integration of WMN units, however few works deals with WMN hardware implementation. In this field our contribution is to conceive and to implements on FPGA a WMN MAC transmitter module. Our implementation, written in hardware description language (HDL) is based on the IEEE 802.11 s standard. The hardware implementation retains a good performance in speed.
△ Less
Submitted 26 April, 2010;
originally announced April 2010.
-
Wireless sensors networks MAC protocols analysis
Authors:
Lamia Chaari,
Lotfi Kamoun
Abstract:
Wireless sensors networks performance are strictly related to the medium access mechanism. An effective one, require non-conventional paradigms for protocol design due to several constraints. An adequate equilibrium between communication improvement and data processing capabilities must be accomplished. To achieve low power operation, several MAC protocols already proposed for WSN. The aim of this…
▽ More
Wireless sensors networks performance are strictly related to the medium access mechanism. An effective one, require non-conventional paradigms for protocol design due to several constraints. An adequate equilibrium between communication improvement and data processing capabilities must be accomplished. To achieve low power operation, several MAC protocols already proposed for WSN. The aim of this paper is to survey and to analyze the most energy efficient MAC protocol in order to categorize them and to compare their performances. Furthermore we have implemented some of WSN MAC protocol under OMNET++ with the purpose to evaluate their performances.
△ Less
Submitted 26 April, 2010;
originally announced April 2010.
