The signal processing (SP) landscape has been enriched by recent advances in artificial intelligence (AI) and machine learning (ML), yielding new tools for signal estimation, classification, prediction, and manipulation. Layered signal representations, nonlinear function approximation and nonlinear signal prediction are now feasible at very large scale in both dimensionality and data size. These are leading to significant performance gains in a variety of long-standing problem domains like speech and Image analysis. As well as providing the ability to construct new classes of nonlinear functions (e.g., fusion, nonlinear filtering). This book will help academics, researchers, developers, graduate and undergraduate students to comprehend complex SP data across a wide range of topical application areas such as social multimedia data collected from social media networks, medical imaging data, data from Covid tests etc. This book focuses on AI utilization in the speech, image, communications and yirtual reality domains.

This book presents various application areas of computing in the automotive sector. The authors explain how computing enhances the performance of vehicles, covering the applications of computing in smart transportation and the future scope. The authors focus on computing for vehicle safety in conjunction with the latest technologies in Internet of Things (IoT). The book provides a holistic approach to computing in an inter-disciplinary and unified view. Topics covered include driverless automated navigation systems, smart transportation, self-learning systems, in-vehicle intelligent systems, and off-road vehicle diagnosis and maintenance, among others. The authors include simulated examples and case studies for better understanding of the technologies and applications. The book is intended for a wide range of readers from students to researchers and industry practitioners and is a useful resource for those planning to pursue research in the area of computing and autonomous driving vehicles.

Electromyography (EMG) signal gives an electrical representation of neuromuscular activation associated with contracting muscle provides information about the performance of muscles and nerves. EMG signal acquires noise while traveling through different tissues. With the appropriate choice of the Wavelet Function (WF), it is possible to remove interference noise. Higher Order Statistics (HOS) can suppress white Gaussian noise in detection, parameter estimation and solve classification problems. Based on the RMS error, it is noticed that WF db2 can perform denoising most effectively among the other WFs (db6, db8, dmey). Power spectrum analysis is performed to the denoised EMG where mean power frequency is calculated to indicate changes in muscle contraction. Gaussianity and linearity tests are conducted to understand changes in muscle contraction. According to the results, increase in muscle contraction provides significant increase in EMG mean power frequency. The study also verifies that the power spectrum of EMG shows a shift to lower frequencies during fatigue. The bispectrum analysis shows that the signal becomes less Gaussian and more linear with increasing muscle force.

Fetal Electrocardiogram (FECG) signal contains potentially precise information that could assist clinicians in making more appropriate and timely decisions during labor. A Back-propagation Neural Network and Adaptive Linear Neural Network have been designed to extract the FECG from the abdominal ECG to assess the fetus during the pregnancy and labor. The neural network was trained to recognize the normal waveform and filtered out the unnecessary artifacts including noises in the ECG signal, including power line interference, motion artifacts, baseline drift, ECG amplitude modulation with respiration and other composite noises. The performance of the designed algorithm for FHR extraction is 93.75%. The algorithm has been modeled using VHDL for hardware modeling of FHR monitoring system, which has been synthesized and fitted into Altera's Stratix II EP2S15F484C3 using the Quartus II version 7.2 Web Edition where the logic and DSP block utilization were 89% and 50% respectively. This research will open up a passage to biomedical researchers and physicians to advocate an excellent understanding of FECG signal and its analysis procedures for FHR monitoring system.

The Set Partitioning in Hierarchical Trees (SPIHT) is modified and a new algorithm is developed, called Modified SPIHT (MSPIHT) using one list to store the co-ordinates of wavelet coefficients. When a coefficient of DWT is found as significant or insignificant, its last error bits will be omitted and rest of the bits will be outputted. MSPIHT is the low memory solution of SPIHT algorithm by eliminating the temporary list LSP and LIP. Absolute zerotree is a good solution of the rapid extension of LIS. The MATLAB simulation result shows that for coding a 512x512, gray-level image, MSPIHT reduce execution time at most 7 times and for decoding at most 11 times at low bit rate, saves at least 0.5625 MBytes of memory. The PSNR value of the reconstructed image using MSPIHT algorithm is reduced by 0.787% with respect to original SPIHT algorithm. This system is implemented on Altera EP2S60F1020C4 FPGA using the software Quartus II 6.0. The results from hardware implementation show that this design has speeded up at most 7129 times faster than that of the results obtained from MATLAB simulations; thereby making it highly promising for real-time and memory limited mobile communication.