Subband Splitting: An Efficient Technique for Solving the Block Permutation Problem in Determined Blind Source Separation
The proposed subband splitting technique effectively solves the block permutation problem in determined blind source separation by splitting the entire frequency range into overlapping subbands and sequentially applying a blind source separation method to each subband.
Orthogonal Mode Decomposition: A Precise and Efficient Approach for Analyzing Finite Discrete Signals
The orthogonal mode decomposition method provides a precise and efficient approach to decompose finite length real signals into a unique and orthogonal set of modes, where each mode is a narrowband function with a clear center frequency and bandwidth.
Efficient Time-Frequency Domain Audio Inpainting Using Autoregressive Modeling
The paper proposes a novel time-frequency domain audio inpainting method, Janssen-TF, which outperforms the recent deep learning-based approach in both objective and subjective evaluations.
Enhancing In-Ear Electrocardiogram Signals Using Denoising Convolutional Autoencoders for Improved Cardiovascular Monitoring
A denoising convolutional autoencoder model can effectively enhance in-ear electrocardiogram signals, improving signal-to-noise ratio, heart rate estimation accuracy, and R-peak detection precision compared to the original noisy in-ear ECG recordings.
Efficient Reconstruction of Periodic Band-Limited Signals from Multi-Input Multi-Output Sampled Data
This paper introduces an efficient FFT-based algorithm to perfectly reconstruct a set of periodic band-limited signals from the samples of the output signals of a multi-input multi-output (MIMO) system.
Zak-OTFS: Pulse Shaping and Its Impact on Predictability and Performance
Gaussian pulse shaping filters in the delay-Doppler domain improve the predictability and performance of Zak-OTFS modulation compared to sinc and root raised cosine filters.
Optimal Signal Detection Using Correlation and Energy Metrics
The optimal detector for distinguishing between a null hypothesis (no signal) and an alternative hypothesis (signal corrupted by noise) is derived by jointly optimizing the transmitted signal and the detector parameters. The detector is based on a linear combination of the correlation and energy of the received signal.
Quantization Analysis and Reconstruction Error Bounds for Bandlimited and Finite-Rate-of-Innovation Signals using Integrate-and-Fire Time Encoding Machines
The quantization step size of the integrate-and-fire time encoding machine (IF-TEM) sampler can be reduced when the maximum frequency of a bandlimited signal or the number of pulses of a finite-rate-of-innovation (FRI) signal is increased. This allows the IF-TEM sampler to achieve a mean squared error (MSE) bound that is roughly 8 dB lower than that of a classical analog-to-digital converter (ADC) with the same number of bits, under specific parameter settings.
Optimal Filter Design for Gear Fault Detection under Time-Varying Speed Conditions using Generalised Envelope Spectrum-Based Signal-to-Noise Objectives
A generalised envelope spectrum-based signal-to-noise objective is proposed to derive optimal filter coefficients that enhance gear fault signatures and attenuate extraneous components under time-varying speed conditions.
Sparse Signal Recovery Guarantees and Cramér-Rao Bounds in the Fractional Fourier Domain
A novel time-domain sparse recovery method that avoids the limitations of transform domain approaches, along with Cramér-Rao Bounds for the sparse parameter estimation problem in the fractional Fourier domain.
© 2024 by Linnk AI