Circuits, Systems and Signal Processing最新文献

This paper proposes a deep hybrid model for stereophonic acoustic echo cancellation (SAEC). A two-stage model is considered, i.e., a deep-learning-based Kalman filter (DeepKalman) and a gated convolutional recurrent network (GCRN)-based postfilter, which are jointly trained in an end-to-end manner. The difference between the proposed DeepKalman filter and the conventional one is twofold. First, the input signal of the DeepKalman filter is a combination of the original two far-end signals and the nonlinear reference signal estimated from the microphone signal directly. Second, a low-complexity recurrent neural network is utilized to estimate the covariance of the process noise, which can enhance the tracking capability of the DeepKalman filter. In the second stage, we adopt GCRN to suppress residual echo and noise by estimating complex masks applied to the output signal of the first stage. Computer simulations confirm the performance advantage of the proposed method over existing SAEC algorithms.

In this paper, the fixed-time synchronization (FXTS) and energy consumption prediction of interconnected memristive neural networks (IMNNs) with discontinuous activation functions are studied. First of all, to deal with discontinuous activation functions, a unified discontinuous controller, which can be applied to other networks, is designed to effectuate FXTS. Based on Lyapunov stability theory, some sufficient conditions for the synchronization of IMNNs are given, and the settling-time is estimated. Meanwhile, to accurately evaluate engineering costs, the energy consumption of the system control process is obtained, and the influence of specific controller parameters on energy consumption is analyzed. Finally, a numerical simulation is presented to verify the correctness of the results.

Image denoising has gained in relevance as a component of image preprocessing due to the increased use of digital images in a range of applications, as well as the degradation of image quality caused by noise introduced by unavoidable occurrences. This work suggests a novel two-stage filter to remove salt and pepper noise from the images. It operates in two stages, the first stage uses an enhanced adaptive type-2 fuzzy noise identifier to identify the corrupted pixel, and the second stage uses a simple linear regression-based approach filter to denoise the corrupted pixel. We first identify a pixel as corrupted or uncorrupted using an improved adaptive type-2 fuzzy-based Gaussian membership function with variables both mean and variance for a specific corrupted image frame. The second step is denoising the damaged pixel using a linear regression-based technique. Herein, we propose a novel co-design method that uses the Gaussian membership function for detection and a linear regression-based denoising technique without any parameter tuning, resulting in better time efficiency. We validate the proposed improved adaptive type-2 fuzzy detection and linear regression-based filter (IAFDLRBF) on a variety of standard images and real-time images with varying noise density. We compare the simulation results with various state-of-the-art methods in terms of various assessment metrics. The results demonstrate the effectiveness of the proposed filter even at high noise densities by providing better detail and edge preservation of an image.

In this paper, the problem of reliable bumpless transfer control for switched systems with actuator faults is investigated. A switching law and a class of reliable controllers via output feedback are constructed such that the bumpless transfer performance and (H_infty ) reliable property of the closed-loop systems are guaranteed. The switching process is divided into two stages, that is, dwell time and state-dependent switching stages. In the dwell time stage, a low bound of the time of the activated subsystem is introduced to restrict the frequent switching and suppress the system bumps. This guarantees the bumpless transfer performance of systems. In the second stage, only the measurable controller states are used to generate the switching signals, which upgrades the feasibility in the frame of output feedback. Finally, to ensure the (H_infty ) reliable property, the proposed controllers are time-varying, which work in the different switching stages. A simulation example is given to demonstrate the effectiveness of the proposed control scheme.

In this paper, the impact of the inductive peaking bandwidth enhancement technique on the input-referred noise performance of a 10 Gbps optical receiver, which includes the CMOS Regulated Cascode (RGC) Transimpedance Amplifier (TIA) is investigated and analyzed. By examining the noise equations in this paper, a novel low-noise design methodology, is introduced. Additionally, through the implementation of the structured ({text{g}}_{text{m}}/{text{I}}_{text{D}}) approach and by selecting the suitable transistor dimensions, the power consumption as well as the noise values are reduced, while the values of the obtained gain and bandwidth are increased. To verify the performance of the designed TIA, the circuit issimulated in HSPICE using 90 nm CMOS technology parameters. The simulation results show the obtained gain value of 55.45 dBΩ, a bandwidth of 7.51 GHz and an input referenced noise value of 15.7 pA/√Hz at the 1.2 V power supply, while the circuit consumes 0.75 mW of power.

Chain-scaling fractional-order memristors (fracmemristors) refers to the concept of implementing them in circuits. Despite its advantages, a few issues that require urgent attention remain, including the effect of input signals on its instantaneous valid frequency range, failure to find the corresponding time-domain electrical characteristic expression of the fracmemristor, and the use of several memristor emulators in circuit schematics. Accordingly, a simple circuit for implementing a chain-scaling fracmemristor (CSF) within a fixed valid frequency range is proposed in this paper. First, a CSF circuit configuration with a fixed valid frequency range is described. Subsequently, a simple circuit schematic of an incremental/decremental CSF is presented, and the corresponding time-domain electrical characteristic expressions of the fracmemristor are obtained. Finally, rich theoretical analysis results are obtained along with the CSF circuit implementation. The accuracy of the theoretical analysis is verified experimentally. In this study, we developed a method to achieve a CSF with a fixed valid frequency range through the replacement of all the capacitors with memcapacitors, which also facilitated the design of circuit schematics.

Accurate measurement is an important prerequisite for sound source localization. In the enclosed environments, noise and reverberation tend to cause localization errors. To address these issues, this paper proposes a compound model particle filter algorithm based on multi-feature. Based on a multi-feature observation, the likelihood function of speaker tracking is constructed for particle filter, and multi-hypothesis and frequency selection function are adopted to establish multi-feature optimization mechanism, including time delay selection and beam output energy fusion. It is found that they effectively solved the difficulty in the simultaneous suppression of noise and reverberation by single feature. Moreover, considering the randomness of speaker motion, a compound model for sound source tracking is developed, where the stability of the speaker tracking system is improved by integrating multi-feature observation into the compound model filtering. The experimental results with both simulated and real acoustic data indicate that the proposed method has better tracking performance, compared with the existing ones with low SNR and strong reverberation as well as highly mobile conditions.

Aiming at the weak robustness of possibilistic fuzzy clustering against noise, a robust possibilistic fuzzy additive partition clustering with master–slave neighborhood information constraints is proposed for high noise image segmentation. This algorithm first constructs a master–slave neighborhood model, which consists of the master neighborhood window of the current pixel and the slave neighborhood window around the master neighborhood pixel. Then, the master–slave neighborhood information is integrated into the possibilistic fuzzy additive partition clustering model, and a novel robust possibilistic fuzzy clustering model incorporating deep local information is constructed. Next, this clustering model is further simplified by Cauchy inequality and a robust master–slave neighborhood information-driven possibilistic fuzzy clustering algorithm is derived by optimization theory. Extensive experimental results indicate that the proposed algorithm is very effective for noisy image segmentation, and its segmentation performance is significantly better than many existing state-of-the-art fuzzy clustering-related algorithms. In short, the work of this paper has profound significance for the development of robust fuzzy clustering theory.

Saturation nonlinearity exists widely in various practical control systems. Modeling and parameter estimation of systems with saturation nonlinearity are of great importance for analyzing their characteristics and controller design. This paper focuses on the identification issue of the input nonlinear Box–Jenkins systems with saturation nonlinearity. The input saturation nonlinearity is presented as a linear parametric expression through the application of a switching function, then the identification model of the system is derived by using the key term separation technique. Based on this model and the data filtering technique, the filtering identification model of the system is given by changing the system structure without changing the relationship between the input and output, which can reduce the interference of the colored noise and improve the identification accuracy. Then a data filtering-based maximum likelihood gradient-based iterative algorithm is proposed to estimate the unknown parameters. The maximum likelihood gradient-based iterative algorithm is provided for comparison. The feasibility and superiority of the proposed approach are emphasized by a simulation example.

In this research paper, we introduce an adaptive block-matching motion estimation algorithm to improve the accuracy and efficiency of motion estimation (ME). First, we present a block generation system that creates blocks of varying sizes based on the detected motion location. Second, we incorporate predictive tools such as early termination and variable window size to optimize our block-matching algorithm. Furthermore, we propose two distinct search patterns to achieve maximum quality and efficiency. We evaluated the proposed algorithms on 20 videos and compared the results with known algorithms, including the full search algorithm (FSA), which is a benchmark for ME accuracy. Our proposed quality-based algorithm shows an improvement of 0.27 dB in peak signal-to-noise ratio (PSNR) on average for reconstructed frames compared to FSA, along with a reduction of 71.66% in searched blocks. Similarly, our proposed efficiency-based method results in a 0.07 dB increase in PSNR and a 97.93% reduction in searched blocks compared to FSA. These findings suggest that our proposed method has the potential to improve the performance of ME in video coding.