In this paper, we develop methods for efficient and accurate information extraction from calcium-imaging-based neural signals. The particular form of information extraction we investigate involves predicting behavior variables linked to animals from which the calcium imaging signals are acquired. More specifically, we develop algorithms to systematically generate compact deep neural network (DNN) models for accurate and efficient calcium-imaging-based predictive modeling. We also develop a software tool, called NeuroGRS, to apply the proposed methods for compact DNN derivation with a high degree of automation. GRS stands for Greedy inter-layer order with Random Selection of intra-layer units, which describes the central algorithm developed in this work for deriving compact DNN structures. Through extensive experiments using NeuroGRS and calcium imaging data, we demonstrate that our methods enable highly streamlined information extraction from calcium images of the brain with minimal loss in accuracy compared to much more computationally expensive approaches.

Applied reconfigurable computing refers to the use of reconfigurable devices to create complex custom computing systems. Examples of the types of devices focused on by the papers in this issue include field-programmable gate arrays (FPGAs) and coarse-grained reconfigurable arrays (CGRAs). Both allow pre-fabricated silicon devices to be configured and reconfigured at run-time to execute highperformance computations. For many important computational areas the resulting systems can perform orders of magnitude faster than software solutions. Two of these computational areas include high-performance signal processing and control engineering as well as neural networks and machine learning. In addition, sophisticated design tools can aid in the creation of such reconfigurable systems and applications, and the development of such tools is an important research area in the field. The papers in this special issue focus on these three areas.

Whilst FPGAs have been used in cloud ecosystems, it is still extremely challenging to achieve high compute density when mapping heterogeneous multi-tasks on shared resources at runtime. This work addresses this by treating the FPGA resource as a service and employing multi-task processing at the high level, design space exploration and static off-line partitioning in order to allow more efficient mapping of heterogeneous tasks onto the FPGA. In addition, a new, comprehensive runtime functional simulator is used to evaluate the effect of various spatial and temporal constraints on both the existing and new approaches when varying system design parameters. A comprehensive suite of real high performance computing tasks was implemented on a Nallatech 385 FPGA card and show that our approach can provide on average 2.9 × and 2.3 × higher system throughput for compute and mixed intensity tasks, while 0.2 × lower for memory intensive tasks due to external memory access latency and bandwidth limitations. The work has been extended by introducing a novel scheduling scheme to enhance temporal utilization of resources when using the proposed approach. Additional results for large queues of mixed intensity tasks (compute and memory) show that the proposed partitioning and scheduling approach can provide higher than 3 × system speedup over previous schemes.

Noise removal in images denotes an interesting and a relatively challenging problem that has captured the attention of many scholars. Recent denoising methods focus on simultaneously restoring noisy images and recovering their semantic features (edges and contours). But preservation of textures, which facilitate interpretation and analysis of complex images, remains an open-ended research question. Classical methods (Total variation and Perona-Malik) and image denoising approaches based on deep neural networks tend to smudge fine details of images. Results from previous studies show that these methods, in addition, can introduce undesirable artifacts into textured images. To address the challenges, we have proposed an image denoising method based on anisotropic diffusion processes. The divergence term of our method contains a diffusion kernel that depends on the evolving image and its gradient magnitude to ensure effective preservation of edges, contours, and textures. Furthermore, a regularization term has been proposed to denoise images corrupted by multiplicative noise. Empirical results demonstrate that the proposed method generates images with higher perceptual and objective qualities.

MIMO system identification is a fundamental concern in a variety of applications. Various iterative and recursive MIMO system identification algorithms exist in the literature. The iterative algorithms suffer from high computational cost due to large matrix computations and recursive algorithms suffer from slow convergence speed. This paper proposes a fast recursive exact least squares algorithm for MIMO system identification with fast convergence speed and low computational cost. The recursions of the algorithm give rise to a lattice structure. The lattice structure is less sensitive to round-off errors, coefficients variations and can also be used for model order reduction. Although in this work we estimate an FIR (finite impulse response) model of the MIMO system, the framework can also be used for IIR (infinite impulse response) models, and for estimating linear periodically time varying (LPTV) and multivariable systems. The theory proposed is validated using simulation results.

This paper presents a reconfigurable frequency response masking (FRM) wideband channelizer architecture which is characterized by low computational and hardware complexity. The proposed hardware efficient architecture is realized by incorporating resource shared non-maximally decimated filter bank in the implementation of the FRM wideband channelizer structure. The coefficients of the proposed architecture are optimized and made multiplier-free using Pareto based meta-heuristic algorithm in the canonic signed digit (CSD) space for reducing the total power consumption of the architecture. The architecture is finally designed and synthesized using Xilinx Vivado and Cadence RTL Encounter compiler for the area and power analysis and is compared with existing channnelizer architectures. The comparison highlights the advantages of the proposed architecture in terms of hardware complexity, power and workload in realizing sharp wideband channel filters.

Fully homomorphic encryption (FHE) is a post-quantum secure cryptographic technology that enables privacy-preserving computing on an untrusted platform without divulging any secret or sensitive information. The core of FHE is the bootstrapping algorithm, which is the intermediate refreshing procedure of a processed ciphertext. However, this step has been the computational bottleneck that prevents real-world deployments among various FHE schemes. This paper, to the best of our knowledge, for the first time, presents a scalable and ultra-highly parallel design for the number theoretic transform (NTT)-based polynomial multiplier with a variable number of reconfigurable processing elements (PEs). Hence, the highest degree of acceleration can be achieved for any targeted hardware platform by implementing as many PEs as possible under the resource constraint. The corresponding addressing and scheduling schemes are also proposed to avoid memory access conflict for the PEs, which yields an extremely high utilization ratio of 99.18% on average. In addition, the latency of the proposed design with the general negative wrapped convolution algorithm is reduced by 59.20% compared to prior works.

To develop an automated pulmonary fibrosis (PF) segmentation methodology using a 3D multi-scale convolutional encoder-decoder approach following the robust atlas-based active volume model in thoracic CT for Rhesus Macaques with radiation-induced lung damage. 152 thoracic computed tomography scans of Rhesus Macaques with radiation-induced lung damage were collected. The 3D input data are randomly augmented with the Gaussian blurring when applying the 3D multi-scale convolutional encoder-decoder (3D MSCED) segmentation method.PF in each scan was manually segmented in which 70% scans were used as training data, 20% scans were used as validation data, and 10% scans were used as testing data. The performance of the method is assessed based on a10-fold cross validation method. The workflow of the proposed method has two parts. First, the compromised lung volume with acute radiation-induced PF was segmented using a robust atlas-based active volume model. Next, a 3D multi-scale convolutional encoder-decoder segmentation method was developed which merged the higher spatial information from low-level features with the high-level object knowledge encoded in upper network layers. It included a bottom-up feed-forward convolutional neural network and a top-down learning mask refinement process. The quantitative results of our segmentation method achieved mean Dice score of (0.769, 0.853), mean accuracy of (0.996, 0.999), and mean relative error of (0.302, 0.512) with 95% confidence interval. The qualitative and quantitative comparisons show that our proposed method can achieve better segmentation accuracy with less variance in testing data. This method was extensively validated in NHP datasets. The results demonstrated that the approach is more robust relative to PF than other methods. It is a general framework which can easily be applied to segmentation other lung lesions.

Grasping the future change trend of soil fertility has great significance in improving the soil quality and achieving high-quality crop production and sustainable agricultural development. However, studies predicting the future change trend of farmland soil fertility are scarce. In this paper, with Yanzhou District of Shandong Province as the research area, a study was conducted based on the sampled data from 2012 to 2017. The data extracted from 2012 to 2016 was used for prediction and that from 2017 was applied for verification. The pH, organic matter, available phosphorus, alkali-hydrolyzed nitrogen and available potassium were selected as indexes of soil fertility. From a socioeconomic perspective, the factors affecting the changes in soil fertility selected in this study include fertilization measures, crop yield, area of arable land, farmers’ income, degree of mechanized operation, irrigated area, pesticide dosage, mulch dosage and rural electricity consumption. Based on this, a stochastic Petri net was used to build a model for predicting the soil fertility change trend. According to the relevant statistical data, the parameters of the model were determined, and by using the solid mathematical basis of the model, the probability of about 0.7852 was calculated out for the soil fertility to decline in the study area in the coming year. By comparing the soil fertility in 2016 and 2017, and further analyzing the changes in soil fertility from 2012 to 2016, the method of predicting the variation trend of soil fertility proposed in this study was verified to be effective.