We propose a convolutional neural network (CNN) architecture for image classification based on subband decomposition of the image using wavelets. The proposed architecture decomposes the input image spectra into multiple critically sampled subbands, extracts features using a single CNN per subband, and finally, performs classification by combining the extracted features using a fully connected layer. Processing each of the subbands by an individual CNN, thereby limiting the learning scope of each CNN to a single subband, imposes a form of structural regularization. This provides better generalization capability as seen by the presented results. The proposed architecture achieves best-in-class performance in terms of total multiply-add-accumulator operations and nearly best-in-class performance in terms of total parameters required, yet it maintains competitive classification performance. We also show the proposed architecture is more robust than the regular full-band CNN to noise caused by weight-and-bias quantization and input quantization.
We describe a new filtering approach in the wavelet domain for image denoising and compression, based on the projections of details subbands coefficients (resultants of the splitting procedure, typical in wavelet domain) onto the approximation subband coefficients (much less noisy). The new algorithm is called Projection Onto Approximation Coefficients (POAC). As a result of this approach, only the approximation subband coefficients and three scalars are stored and/or transmitted to the channel. Besides, with the elimination of the details subbands coefficients, we obtain a bigger compression …
Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院）；Korea Aerospace Research Institute (KARI)（韩国航空宇宙研究院）
Multi-spectral satellite imaging sensors acquire various spectral band images such as red (R), green (G), blue (B), near-infrared (N), etc. Thanks to the unique spectroscopic property of each spectral band with respective to the objects on the ground, multi-spectral satellite imagery can be used for various geological survey applications. Unfortunately, image artifacts from imaging sensor noises often affect the quality of scenes and have negative impacts on the applications of satellite imagery. Recently, deep learning approaches have been extensively explored for the removal of noises in sat…
Chittagong University of Engineering and Technology（吉大港工程与技术大学）；University of Malaya（马来亚大学）；American International University-Bangladesh（美国国际大学-孟加拉）
Digital watermarking is the process to hide digital pattern directly into a digital content. Digital watermarking techniques are used to address digital rights management, protect information and conceal secrets. An invisible non-blind watermarking approach for gray scale images is proposed in this paper. The host image is decomposed into 3-levels using Discrete Wavelet Transform. Based on the parent-child relationship between the wavelet coefficients the Set Partitioning in Hierarchical Trees (SPIHT) compression algorithm is performed on the LH3, LH2, HL3 and HL2 subbands to find out the sign…
In this paper, a new probability density function (pdf) is proposed to model the statistics of wavelet coefficients, and a simple Kalman's filter is derived from the new pdf using Bayesian estimation theory. Specifically, we decompose the speckled image into wavelet subbands, we apply the Kalman's filter to the high subbands, and reconstruct a despeckled image from the modified detail coefficients. Experimental results demonstrate that our method compares favorably to several other despeckling methods on test synthetic aperture radar (SAR) images.
The bilateral filter is a useful nonlinear filter which without smoothing edges, it does spatial averaging. In the literature, the effectiveness of this method for image denoising is shown. In this paper, an extension of this method is proposed which is based on complex wavelet transform. In fact, the bilateral filtering is applied to the low-frequency (approximation) subbands of the decomposed image using complex wavelet transform, while the thresholding approach is applied to the high frequency subbands. Using the bilateral filter in the complex wavelet domain forms a new image denoising fra…
Prior studies using graph neural networks (GNNs) for image classification have focused on graphs generated from a regular grid of pixels or similar-sized superpixels. In the latter, a single target number of superpixels is defined for an entire dataset irrespective of differences across images and their intrinsic multiscale structure. On the contrary, this study investigates image classification using graphs generated from an image-specific number of multiscale superpixels. We propose WaveMesh, a new wavelet-based superpixeling algorithm, where the number and sizes of superpixels in an image a…
Shaanxi Normal University（陕西师范大学）；Nanjing University of Science and Technology（南京理工大学）；Harbin Institute of Technology（哈尔滨工业大学）
Recent years have witnessed the great success of deep convolutional neural networks (CNNs) in image denoising. Albeit deeper network and larger model capacity generally benefit performance, it remains a challenging practical issue to train a very deep image denoising network. Using multilevel wavelet-CNN (MWCNN) as an example, we empirically find that the denoising performance cannot be significantly improved by either increasing wavelet decomposition levels or increasing convolution layers within each level. To cope with this issue, this paper presents a multi-level wavelet residual network (…
We describe a new method for superresolution of still images (in the wavelet domain) based on the reconstruction of missing details subbands pixels at a given ith level via Rule of Three (Ro3) between pixels of approximation subband of such level, and pixels of approximation and detail subbands of (i+1)th level. The histogramic profiles demonstrate that Ro3 is the appropriate mechanism to recover missing detail subband pixels in these cases. Besides, with the elimination of the details subbands pixels (in an eventual compression scheme), we obtain a bigger compression rate. Experimental result…
SAOCOM Mission, National Commission of Space Activities (CONAE)（阿根廷国家空间活动委员会 SAOCOM 任务）
This paper describes a novel projection algorithm, the Projection Onto Span Algorithm (POSA) for wavelet-based superresolution and removing speckle (in wavelet domain) of unknown variance from Synthetic Aperture Radar (SAR) images. Although the POSA is good as a new superresolution algorithm for image enhancement, image metrology and biometric identification, here one will use it like a tool of despeckling, being the first time that an algorithm of super-resolution is used for despeckling of SAR images. Specifically, the speckled SAR image is decomposed into wavelet subbands, POSA is applied t…
We describe a novel method for removing speckle (in wavelet domain) of unknown variance from SAR images. The me-thod is based on the following procedure: We apply 1) Bidimentional Discrete Wavelet Transform (DWT-2D) to the speckled image, 2) scaling and rounding to the coefficients of the highest subbands (to obtain integer and positive coefficients), 3) bit-slicing to the new highest subbands (to obtain bit-planes), 4) then we apply the Systholic Boolean Orthonormalizer Network (SBON) to the input bit-plane set and we obtain two orthonormal output bit-plane sets (in a Boolean sense), we proje…
ChWDTA: Channel-wise Wavelet-Domain Transformer Attention and Entropy Modeling for Learned Image Compression
Electrical and Computer Engineering Department, The University of British Columbia（英属哥伦比亚大学电气与计算机工程系）；School of Engineering Science, Simon Fraser University（西蒙弗雷泽大学工程科学学院）；School of Electronic Science and Technology, Eastern Institute of Technology（电子科学与技术学院，东部技术学院）；Google LLC（谷歌公司）
With the heterogeneous nature of tissue texture, using a single resolution approach for optimum classification might not suffice. In contrast, a multiresolution wavelet packet analysis can decompose the input signal into a set of frequency subbands giving the opportunity to characterise the texture at the appropriate frequency channel. An adaptive best bases algorithm for optimal bases selection for meningioma histopathological images is proposed, via applying the fractal dimension (FD) as the bases selection criterion in a tree-structured manner. Thereby, the most significant subband that bet…
Convolutional neural networks (CNNs) have been applied to learn spatial features for high-resolution (HR) synthetic aperture radar (SAR) image classification. However, there has been little work on integrating the unique statistical distributions of SAR images which can reveal physical properties of terrain objects, into CNNs in a supervised feature learning framework. To address this problem, a novel end-to-end supervised classification method is proposed for HR SAR images by considering both spatial context and statistical features. First, to extract more effective spatial features from SAR …
We propose a novel motion estimation/compensation (ME/MC) method for wavelet-based (in-band) motion compensated temporal filtering (MCTF), with application to low-bitrate video coding. Unlike the conventional in-band MCTF algorithms, which require redundancy to overcome the shift-variance problem of critically sampled (complete) discrete wavelet transforms (DWT), we perform ME/MC steps directly on DWT coefficients by avoiding the need of shift-invariance. We omit upsampling, the inverse-DWT (IDWT), and the calculation of redundant DWT coefficients, while achieving arbitrary subpixel accuracy w…
WaveLLM: Wavelet-Domain LLM Entropy Modeling for Lossless Image Compression
University of Science and Technology of China（中国科学技术大学）；Nanyang Technological University（南洋理工大学）；Institute of Artificial Intelligence, Hefei Comprehensive National Science Center（合肥综合性国家科学中心人工智能研究院）
National Yang Ming Chiao Tung University（阳明交通大学）；University at Albany - SUNY（奥尔巴尼大学 - 纽约州立大学）；University at Buffalo（布法罗大学）
Pooling is a simple but essential layer in modern deep CNN architectures for feature aggregation and extraction. Typical CNN design focuses on the conv layers and activation functions, while leaving the pooling layers with fewer options. We introduce the Learning Discrete Wavelet Pooling (LDW-Pooling) that can be applied universally to replace standard pooling operations to better extract features with improved accuracy and efficiency. Motivated from the wavelet theory, we adopt the low-pass (L) and high-pass (H) filters horizontally and vertically for pooling on a 2D feature map. Feature sign…
University of Isfahan（伊斯法罕大学）；Hunan University（湖南大学）；Isfahan University of Medical Sciences（伊斯法罕医科大学）；Sichuan University（四川大学）
The data-driven sparse methods such as synthesis dictionary learning (e.g., K-SVD) and sparsifying transform learning have been proven effective in image denoising. However, they are intrinsically single-scale which can lead to suboptimal results. We propose two methods developed based on wavelet subbands mixing to efficiently combine the merits of both single and multiscale methods. We show that an efficient multiscale method can be devised without the need for denoising detail subbands which substantially reduces the runtime. The proposed methods are initially derived within the framework of…
The application of wavelet transforms to Synthetic Aperture Radar (SAR) imagery has improved despeckling performance. To deduce the problem of filtering the multiplicative noise to the case of an additive noise, the wavelet decomposition is performed on the logarithm of the image gray levels. The detail coefficients produced by the bidimensional discrete wavelet transform (DWT-2D) needs to be thresholded to extract out the speckle in highest subbands. An initial threshold value is estimated according to the noise variance. In this paper, an additional fuzzy thresholding approach for automatic …
Huazhong University of Science and Technology（华中科技大学）
Rain removal from a single image is a challenge which has been studied for a long time. In this paper, a novel convolutional neural network based on wavelet and dark channel is proposed. On one hand, we think that rain streaks correspond to high frequency component of the image. Therefore, haar wavelet transform is a good choice to separate the rain streaks and background to some extent. More specifically, the LL subband of a rain image is more inclined to express the background information, while LH, HL, HH subband tend to represent the rain streaks and the edges. On the other hand, the accum…
University of California, San Diego（加州大学圣地亚哥分校）；University of Dayton（代顿大学）
We propose a novel lossless and lossy compression scheme for color filter array~(CFA) sampled images based on the wavelet transform of them. Our analysis suggests that the wavelet coefficients of HL and LH subbands are highly correlated. Hence, we decorrelate Mallat wavelet packet decomposition to further sparsify the coefficients. In addition, we develop a camera processing pipeline for compressing CFA sampled images aimed at maximizing the quality of the color images constructed from the compressed CFA sampled images. We validated our theoretical analysis and the performance of the proposed …
Middle East Technical University（中东技术大学）；Tohoku University（东北大学）
We propose a new framework, called Hierarchical Multi-resolution Mesh Networks (HMMNs), which establishes a set of brain networks at multiple time resolutions of fMRI signal to represent the underlying cognitive process. The suggested framework, first, decomposes the fMRI signal into various frequency subbands using wavelet transforms. Then, a brain network, called mesh network, is formed at each subband by ensembling a set of local meshes. The locality around each anatomic region is defined with respect to a neighborhood system based on functional connectivity. The arc weights of a mesh are e…
WaveSFNet: A Wavelet-Based Codec and Spatial--Frequency Dual-Domain Gating Network for Spatiotemporal Prediction
College of Software Engineering, Sichuan University（四川大学软件工程学院）；College of Computer Science, Sichuan University（四川大学计算机科学学院）；Sichuan University National Key Laboratory of Fundamental Science on Synthetic Vision（四川大学合成视觉基础科学国家重点实验室）
Frequency spectrum has played a significant role in learning unique and discriminating features for object recognition. Both low and high frequency information present in images have been extracted and learnt by a host of representation learning techniques, including deep learning. Inspired by this observation, we introduce a novel class of adversarial attacks, namely `WaveTransform', that creates adversarial noise corresponding to low-frequency and high-frequency subbands, separately (or in combination). The frequency subbands are analyzed using wavelet decomposition; the subbands are corrupt…
WaveDH: Wavelet Sub-bands Guided ConvNet for Efficient Image Dehazing
Artificial Intelligence Graduate School, Gwangju Institute of Science and Technology (GIST)（光州科学技术院人工智能研究生院）；School of Electronic Engineering, Xidian University（西安电子科技大学电子工程学院）；School of Electrical Engineering and Computer Science, Gwangju Institute of Science and Technology (GIST)（光州科学技术院电气与计算机工程学院）
Image splicing detection is of fundamental importance in digital forensics and therefore has attracted increasing attention recently. In this paper, a color image splicing detection approach is proposed based on Markov transition probability of quaternion component separation in quaternion discrete cosine transform (QDCT) domain and quaternion wavelet transform (QWT) domain. Firstly, Markov features of the intra-block and inter-block between block QDCT coefficients are obtained from the real part and three imaginary parts of QDCT coefficients respectively. Then, additional Markov features are …
RatioWaveNet: A Learnable RDWT Front-End for Robust and Interpretable EEG Motor-Imagery Classification
Department of Electrical, Electronics and Informatics Engineering, University of Catania（电子、电子与信息工程系，卡塔尼亚大学）；Department of Engineering, University of Palermo（工程系，巴勒莫大学）
Unlocking Fine-Grained Details with Wavelet-based High-Frequency Enhancement in Transformers
RWTH Aachen University（亚琛工业大学）；Iran University of Science and Technology（伊朗科技大学）；Universiti Kebangsaan（马来西亚国民大学）；University of Regensburg（雷根斯堡大学）；Shahid Beheshti University（沙希德·贝赫什提大学）
GDSR: Global-Detail Integration through Dual-Branch Network with Wavelet Losses for Remote Sensing Image Super-Resolution
School of Computer Technology and Application, Qinghai University（青海大学计算机技术与应用学院）；Qinghai Provincial Laboratory for Intelligent Computing and Application, Qinghai University（青海省智能计算与应用实验室）；Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）；School of Computer and Information Science, Qinghai Institute of Technology（青海理工大学计算机与信息科学学院）；College of Agriculture and Animal Husbandry, Qinghai University（青海大学农业与动物科学学院）
Image denoising is of vital importance in many imaging or computer vision related areas. With the convolutional neural networks showing strong capability in computer vision tasks, the performance of image denoising has also been brought up by CNN based methods. Though CNN based image denoisers show promising results on this task, most of the current CNN based methods try to learn the mapping from noisy image to clean image directly, which lacks the explicit exploration of prior knowledge of images and noises. Natural images are observed to obey the reciprocal power law, implying the low-freque…
