Improving performance of deep learning models with axiomatic attribution priors and expected gradients
Recent research has demonstrated that feature attribution methods for deep networks can
themselves be incorporated into training; these attribution priors optimize for a model whose …
themselves be incorporated into training; these attribution priors optimize for a model whose …
A novel strategy for signal denoising using reweighted SVD and its applications to weak fault feature enhancement of rotating machinery
Singular value decomposition (SVD), as an effective signal denoising tool, has been
attracting considerable attention in recent years. The basic idea behind SVD denoising is to …
attracting considerable attention in recent years. The basic idea behind SVD denoising is to …
Improvement of kurtosis-guided-grams via Gini index for bearing fault feature identification
A group of kurtosis-guided-grams, such as Kurtogram, Protrugram and SKRgram, is
designed to detect the resonance band excited by faults based on the sparsity index …
designed to detect the resonance band excited by faults based on the sparsity index …
Roadmap on holography
JT Sheridan, RK Kostuk, AF Gil, Y Wang, W Lu… - Journal of …, 2020 - iopscience.iop.org
From its inception holography has proven an extremely productive and attractive area of
research. While specific technical applications give rise to 'hot topics', and three-dimensional …
research. While specific technical applications give rise to 'hot topics', and three-dimensional …
Noise robust face hallucination via locality-constrained representation
Recently, position-patch based approaches have been proposed to replace the probabilistic
graph-based or manifold learning-based models for face hallucination. In order to obtain the …
graph-based or manifold learning-based models for face hallucination. In order to obtain the …
Generalized Gini indices: Complementary sparsity measures to Box-Cox sparsity measures for machine condition monitoring
Sparsity measures that can quantify the sparsity of signals are often used as objective
functions of signal processing and machine learning algorithms (eg, sparse filtering …
functions of signal processing and machine learning algorithms (eg, sparse filtering …
Health assessment of rotating machinery using a rotary encoder
M Zhao, J Lin - IEEE Transactions on Industrial Electronics, 2017 - ieeexplore.ieee.org
In this paper, a systematic framework is established for health assessment of rotating
machinery using a rotary encoder. In this framework, spectral quadratic weighting is first …
machinery using a rotary encoder. In this framework, spectral quadratic weighting is first …
Learning explainable models using attribution priors
Two important topics in deep learning both involve incorporating humans into the modeling
process: Model priors transfer information from humans to a model by regularizing the …
process: Model priors transfer information from humans to a model by regularizing the …
Refocusing criterion via sparsity measurements in digital holography
Several automatic approaches have been proposed in the past to compute the refocus
distance in digital holography (DH). However most of them are based on a maximization or …
distance in digital holography (DH). However most of them are based on a maximization or …
Application of an improved MCKDA for fault detection of wind turbine gear based on encoder signal
Due to severe working condition, unexpected failures in wind turbine gearbox become
rather frequent and may lead to long downtime or even catastrophic casualties. However …
rather frequent and may lead to long downtime or even catastrophic casualties. However …