Brain analysis mainly relies on complex recording techniques and on advanced signal processing tools used to interpret these recordings. In neurophysiological time series, as strong neural oscillations are observed in...
详细信息
Brain analysis mainly relies on complex recording techniques and on advanced signal processing tools used to interpret these recordings. In neurophysiological time series, as strong neural oscillations are observed in the mammalian brain, the natural processing tools are centered on spectral analysis, Fourier decomposition, and on linear filtering into canonical frequency bands. While this approach has had significant impact in neuroscience, it may give a misleading representation of the signal. Indeed, it is standard to see neuro-scientists consider small subsets of coefficients, implicitly assuming that the signals of interest are narrow-band, which turns out to be too reductive. Multiple warnings have been raised about this fallacy, and about the need of more appropriate methods to represent the signals. More generally, a large number of neuroscientific studies use ad-hoc recipes to analyze time series and describe their properties. Importantly, these methods are heavily based on narrow-band filtering and on custom correlation metrics, and they fail to give a goodness of fit. Therefore, setting the parameters of these methods can only be driven by how much they lead to a strong value of the metric. As a consequence, even though these metrics give reasonable information, a legitimate and controlled comparison of methods and parameters, and therefore of the results, is impossible. This is the case for instance for a phenomenon known as phase-amplitude coupling (PAC), which consists in an amplitude modulation of a high frequency signal, time-locked with a specific phase of a slow neural oscillation. In this work, we first propose to use driven autoregressive models (DAR) on neuro- physiological time-series. These models give a spectral representation of the signal conditionally to another signal, and thus are able to capture PAC in a probabilistic model of the signal, for which statistical inference is fast and well-posed. Giving a proper model to the signal enables
Simultaneous sparse approximation (SSA) seeks to represent a set of dependent signals using sparse vectors with identical supports. The SSA model has been used in various signal and image processing applications invol...
详细信息
ISBN:
(纸本)9781665459068
Simultaneous sparse approximation (SSA) seeks to represent a set of dependent signals using sparse vectors with identical supports. The SSA model has been used in various signal and image processing applications involving multiple correlated input signals. In this paper, we propose algorithms for convolutional SSA (CSSA) based on the alternating direction method of multipliers. Specifically, we address the CSSA problem with different sparsity structures and the convolutional feature learning problem in multimodal data/signals based on the SSA model. We evaluate the proposed algorithms by applying them to multimodal and multifocus image fusion problems.
Priors play an important role of regularizers in image deblurring algorithms. Image priors are frequently studied and many forms were proposed in the literature. Blur priors are considered less important and the most ...
详细信息
ISBN:
(纸本)9781728163956
Priors play an important role of regularizers in image deblurring algorithms. Image priors are frequently studied and many forms were proposed in the literature. Blur priors are considered less important and the most common forms are simple uniform distributions with domain constraints. We propose a more informative blur prior based on the notion of atomic norm which favors blurs composed of line segments and is suitable for motion blur. The prior is formulated as a linear program that can be inserted into any optimization task. Evaluation is conducted on blind deblurring of moving objects.
We address the problem of efficient convolutional sparse coding (CSC) and develop a non-convex-penalty-regularized CSC formulation, namely, minimax-concave CSC ((MCSC)-S-2). (MCSC)-S-2 leads to an optimal sparse repre...
详细信息
ISBN:
(纸本)9781509066315
We address the problem of efficient convolutional sparse coding (CSC) and develop a non-convex-penalty-regularized CSC formulation, namely, minimax-concave CSC ((MCSC)-S-2). (MCSC)-S-2 leads to an optimal sparse representation than the standard l(1)-penalty based approach. In addition, suitable convergence guarantees can also be provided for (MCSC)-S-2. We propose a convolutional iterative firmthresholding algorithm (CIFTA) building on our previously proposed IFTA, and its deep-unfolded version, namely, convolutionalFirmNet (ConFirmNet). As an application, we develop the ConFirmNet based sparse autoencoder (ConFirmNet-SAE) for learning an application-specific convolutional dictionary, the applications being image denoising and inpainting. Further, we also show that training ConFirmNet-SAE with the Huber loss imparts robustness to outliers. It also turns out that ConFirmNet-SAE is robust to mismatch between training and test noise conditions than convolutional learned iterative soft-thresholding algorithm (LISTA).
Porosity defects can be found in many engineering structures and their inspection remains a challenge in the field of ultrasonic non-destructive testing. In this paper, ultrasonic array imaging of porosity defects has...
详细信息
Porosity defects can be found in many engineering structures and their inspection remains a challenge in the field of ultrasonic non-destructive testing. In this paper, ultrasonic array imaging of porosity defects has been studied with the aim of improving the image quality in the "dead zone", which is caused by the masking effects of the uppermost pores. The proposed approach first extracts contributions of the uppermost pores based on a single scattering model by adopting convolutional sparse coding. The extracted dominant contributions are then subtracted from the array data before forming an image, facilitating detection and localization of pores in the shadow zone. The performance of the proposed approach has been studied in simulation and experiments, and the mean localization errors of the pores are small (i.e., within 0.27 mm or 0.14 ������ ). In addition, the effects of measurement noise and imaging parameters on robustness of the imaging result have been analyzed, providing guidelines for practical implementation of the proposed approach.
Multispectral and hyperspectral image fusion (MS/HS fusion) aims to generate a high-resolution hyperspectral (HRHS) image by fusing a high-resolution multispectral (HRMS) and a low-resolution hyperspectral (LRHS) imag...
详细信息
Multispectral and hyperspectral image fusion (MS/HS fusion) aims to generate a high-resolution hyperspectral (HRHS) image by fusing a high-resolution multispectral (HRMS) and a low-resolution hyperspectral (LRHS) images. The deep unfolding-based MS/HS fusion method is a representative deep learning paradigm due to its excellent performance and sufficient interpretability. However, existing deep unfolding-based MS/HS fusion methods only rely on a fixed linear degradation model, which focuses on modeling the relationships between HRHS and HRMS, as well as HRHS and LRHS. In this paper, we break free from this observation model framework and propose a new observation model. Firstly, the proposed observation model is built based on the convolutional sparse coding (CSC) technique, and then a proximal gradient algorithm is designed to solve this model. Secondly, we unfold the iterative algorithm into a deep network, dubbed as MHF-CSCNet, where the proximal operators are learned using convolutional neural networks. Finally, all trainable parameters can be automatically learned end-to-end from the training pairs. Experimental evaluations conducted on various benchmark datasets demonstrate the superiority of our method both quantitatively and qualitatively compared to other state-of-the-art methods.
convolutional analysis operator learning, which takes advantage of the ability to extract and store overlapping blocks across training signals, has been the subject of much research in computer vision applications. Th...
详细信息
convolutional analysis operator learning, which takes advantage of the ability to extract and store overlapping blocks across training signals, has been the subject of much research in computer vision applications. The redundant filter learned by this method has the advantages of both constraining orthogonality and promoting diversity. This study, therefore, applies the convolution analysis operator to the field of image fusion and proposes a multimodal image-fusion method based on the convolution analysis operator. Experimental results show that this method performs better than the comparison methods as it not only retains the edges in the reconstructed image, but also considers the global structure of the image.
While convolutionalsparse representations enjoy a number of useful properties, they have received limited attention for image reconstruction problems. The present paper compares the performance of block-based and con...
详细信息
ISBN:
(纸本)9781538646595
While convolutionalsparse representations enjoy a number of useful properties, they have received limited attention for image reconstruction problems. The present paper compares the performance of block-based and convolutionalsparse representations in the removal of Gaussian white noise. The usual formulation of the convolutional sparse coding problem is slightly inferior to the block-based representations in this problem, but the performance of the convolutional form can be boosted beyond that of the block-based form by the inclusion of suitable penalties on the gradients of the coefficient maps.
Objects in fine-grained categories always share a high degree of shape similarity, making both "localizing discriminative parts" and "learning appearance descriptors" extremely difficult. We propos...
详细信息
ISBN:
(纸本)9781467399623
Objects in fine-grained categories always share a high degree of shape similarity, making both "localizing discriminative parts" and "learning appearance descriptors" extremely difficult. We propose a framework to leverage 2D+3D cues to handle above two challenges. Towards the goal of image alignment to localize discriminative parts, traditional methods rely on either manual part annotation or image segmentation. Instead, our framework leverages each image's 3D camera pose estimation to align images;Towards the goal of "learning appearance descriptors" confined with small training data and memory/computation cost, we propose an unsupervised convolutional sparse coding (CSC) + manifold learning that significantly reduces model complexity, but still successfully produces highly diverse feature filters like deep neural network. Our experimental results attest the advocated framework's accuracy is comparable to a deep network, demonstrating its great potential on mobile devices.
暂无评论