Low-light Image Enhancement (LIE) aims at improving contrast and restoring details for images captured in lowlight conditions. Most of the previous LIE algorithms adjust illumination using a single input image with several handcrafted priors. Those solutions, however, often fail revealing due to limited information poor adaptability To this end, we propose PairLIE, an unsupervised approach that learns adaptive priors from low-light pairs. First, network is expected generate same clean as two...

Synchrosqueezing transform (SST) is a currently proposed novel postprocessing time-frequency (TF) analysis tool. It has been widely shown that SST able to improve TF representation. However, so far, how the resolution while ensuring accuracy of signal reconstruction still an open question, particularly for vibration with time-varying instantaneous frequency (IF) characteristics, due fact signals mechanical equipment usually contain many types noise generated by harsh operating conditions,...

Time-frequency analysis (TFA) is considered as a useful tool to extract the time-variant features of nonstationary signal. In this paper, new method called demodulated high-order synchrosqueezing transform (DHST) proposed. The DHST introduces two-step algorithm, namely, and achieve compact time-frequency representation (TFR) while enabling reconstruction signal from TFR. performance proposed in paper validated by both simulated experimental signals including bat echolocation vibration...

The rolling element bearing is easy to be malfunctioning due the harsh operation. When a fault exists in bearing, it can generate periodical or quasi-periodical impulses, which are important features for detection. These impulses may submerged background noise and interferences of other unrelated components. spectral kurtosis, its fast realization, kurtogram, have been widely used diagnosis by extracting impulsive feature. However, performance weakened fixed decomposition scheme prior...

Time-frequency analysis (TFA) is regarded as an efficient technique to reveal the hidden characteristics of oscillatory signal. At present, traditional TFA methods always construct signal model in time domain and assume instantaneous features modes be continuous. Thus, most these approaches fail tackle some specific kinds impulselike signal, including shock vibration waves, damped tones, or marine mammals. This article introduces a new method called generalized horizontal synchrosqueezing...

Recent generalizable fault diagnosis researches have effectively tackled the distributional shift between unseen working conditions. Most of them mainly focus on learning domain-invariant representation through feature-level methods. However, increasing numbers domains may lead to features contain instance-level spurious correlations, which impact previous models' ability. To address limitations, we propose Fourier-based Augmentation Reconstruction Network, namely FARNet.The methods are...

Anomaly detection (AD) in 3D point clouds is crucial a wide range of industrial applications, especially various forms precision manufacturing. Considering the demand for reliable AD, several methods have been developed. However, most these approaches typically require training separate models each category, which memory-intensive and lacks flexibility. In this paper, we propose novel Point-Language model with dual-prompts ANomaly dEtection (PLANE). The approach leverages multi-modal prompts...

Pan-sharpening aims to preserve the spectral information of multi-spectral (MS) image while leveraging high-frequency details from guided high-resolution panchromatic (PAN) enhance its spatial resolution. The key challenge is how MS and PAN as much possible. Diffusion models have achieved favorable results in restoration synthesis tasks but suffer excessive computational resource time consumption. In this paper, we design a novel computationally efficient diffusion-based pan-sharpening...

Hyperspectral image (HSI) reconstruction aims to restore the original 3D HSIs from 2D hyperspectral snapshot compressive images (SCIs). The key high-fidelity HSI lies in designing refined spatial and spectral attention mechanisms, which are crucial for generating fine-grained representations of based on limited information available SCI. Recently, Mamba has demonstrated remarkable performance efficiency modeling correlations. Its implicit mechanism generates three orders magnitude more...

Low-light image enhancement (LIE) aims at precisely and efficiently recovering an degraded in poor illumination environments. Recent advanced LIE techniques are using deep neural networks, which require lots of low-normal light pairs, network parameters, computational resources. As a result, their practicality is limited. In this work, we devise novel unsupervised framework based on diffusion priors lookup tables (DPLUT) to achieve efficient low-light recovery. The proposed approach...

Wind turbines usually operate under nonstationary conditions, such as wide-range speed fluctuation and time-varying load. Its critical component, the planetary gearbox, is prone to malfunction or failure, which leads downtime repair costs. Therefore, fault diagnosis condition monitoring for gearbox in wind a vital research topic. Meanwhile, signals measured by vibration sensors mounted exhibit features. In this study, novel time-frequency method based on high-order synchrosqueezing transform...

As one of the most important and essential elements machines, rolling element bearings always fail due to severe operating environment. Bearing failures usually result in periodic impulses, which are crucial feature for bearing diagnosis. These impulses may be overwhelmed by background noises or other unrelated components. Many traditional features time domain such as kurtosis root mean square (rms) invalid some cases. They ineffective detecting impulses. This paper proposed a novel...

The empirical wavelet transform (EWT) has shown its effectiveness in some applications. However, when noisy and nonstationary signals are analyzed, local maxima may appear be retained the peak sequence mistakenly, so improper segmentation frequency domain will occur. In our research, morphological EWT (MEWT) method is proposed based on filters (MFs) 1-D Otsu to mitigate boundary drawback of EWT, it can applied chatter detection because good performance finding optimal band. First,...

This article introduces the second-order transient-extracting transform (TET2) to extract transient components from a nonstationary signal. Different traditional (TET1), proposed method is based on more general frequency-domain signal model, termed Gaussian-modulated linear group delay model. The first step of computation (GD) estimator that describes temporal positioning ridges in time-frequency (TF) plane. Based this GD estimator, we can then obtain an energy-concentrated TF...