A large amount of data collected from sensors exhibits Gaussian noise characteristics, making denoising and related processing critical. However, data scarcity can lead to overfitting, posing challenges in training de...
详细信息
A large amount of data collected from sensors exhibits Gaussian noise characteristics, making denoising and related processing critical. However, data scarcity can lead to overfitting, posing challenges in training deep learning-based denoising methods. While various data augmentation methods have been proposed, they do not provide a means to augment original data to large-scale data while preserving the exact noise distribution. To address this, we introduce a novel data augmentation method for data with additive white Gaussian noise (AWGN). Our method is based on two main premises: first, orthogonal transforms preserve the probability distribution of AWGN;second, the signals we aim to recover generally exhibit smooth characteristics, unlike noise. Building on these premises, we propose adaptive smoothness-promoting orthogonal transforms for augmenting limited existing data. We evaluated the proposed method in Gaussian denoising tasks with limited data and confirmed that it achieves substantial improvement in deep learning model performance, comparable to those obtained with sufficient data.
The images captured under improper exposure conditions often suffer from contrast degradation and detail distortion. Contrast degradation will destroy the statistical properties of low-frequency components, while deta...
详细信息
ISBN:
(纸本)9798350301298
The images captured under improper exposure conditions often suffer from contrast degradation and detail distortion. Contrast degradation will destroy the statistical properties of low-frequency components, while detail distortion will disturb the structural properties of high-frequency components, leading to the low-frequency and high-frequency components being mixed and inseparable. This will limit the statistical and structural modeling capacity for exposure correction. To address this issue, this paper proposes to decouple the contrast enhancement and detail restoration within each convolution process. It is based on the observation that, in the local regions covered by convolution kernels, the feature response of low-/high-frequency can be decoupled by addition/difference operation. To this end, we inject the addition/difference operation into the convolution process and devise a Contrast Aware (CA) unit and a Detail Aware (DA) unit to facilitate the statistical and structural regularities modeling. The proposed CA and DA can be plugged into existing CNN-based exposure correction networks to substitute the Traditional Convolution (TConv) to improve the performance. Furthermore, to maintain the computational costs of the network without changing, we aggregate two units into a single TConv kernel using structural re-parameterization. Evaluations of nine methods and five benchmark datasets demonstrate that our proposed method can comprehensively improve the performance of existing methods without introducing extra computational costs compared with the original networks. The codes will be publicly available.
Few-shot image generation is a challenging task even using the state-of-the-art Generative Adversarial Networks (GANs). Due to the unstable GAN training process and the limited training data, the generated images are ...
详细信息
ISBN:
(数字)9781665469463
ISBN:
(纸本)9781665469463
Few-shot image generation is a challenging task even using the state-of-the-art Generative Adversarial Networks (GANs). Due to the unstable GAN training process and the limited training data, the generated images are often of low quality and low diversity. In this work, we propose a new "editing-based" method, i.e., Attribute Group Editing (AGE), for few-shot image generation. The basic assumption is that any image is a collection of attributes and the editing direction for a specific attribute is shared across all categories. AGE examines the internal representation learned in GANs and identifies semantically meaningful directions. Specifically, the class embedding, i.e., the mean vector of the latent codes from a specific category, is used to represent the category-relevant attributes, and the category-irrelevant attributes are learned globally by Sparse Dictionary Learning on the difference between the sample embedding and the class embedding. Given a GAN well trained on seen categories, diverse images of unseen categories can be synthesized through editing category-irrelevant attributes while keeping category-relevant attributes unchanged. Without re-training the GAN, AGE is capable of not only producing more realistic and diverse images for downstream visual applications with limited data but achieving controllable image editing with interpretable category-irrelevant directions. Code is available at https://***/UniBester/AGE.
Digital libraries generally need to process a large volume of diverse document types. The collection and tagging of metadata is a long, error-prone, workforce-consuming task. We are attempting to build an automatic me...
详细信息
ISBN:
(纸本)9781665417709
Digital libraries generally need to process a large volume of diverse document types. The collection and tagging of metadata is a long, error-prone, workforce-consuming task. We are attempting to build an automatic metadata extractor for digital libraries. In this work, we present the Heterogeneous Learning Resources (HLR) dataset for document image classification. The individual learning resource is first decomposed into its constituent document images (sheets) which are then passed through an OCR tool to obtain the textual representation. The document image and its textual content are classified with state-of-the-art classifiers. Finally, the labels of the constituent document images are used to predict the label of the overall document.
image edge detection is a fundamental process in computer vision. image edges represent the major fraction of information in an image. Traditional edge-detection techniques focus on the gradient calculation method. In...
详细信息
ISBN:
(纸本)9781479921867
image edge detection is a fundamental process in computer vision. image edges represent the major fraction of information in an image. Traditional edge-detection techniques focus on the gradient calculation method. In this paper, for the first time, the statistical patternrecognition method is used to detect the edge after the real-time image was processed via the median filtering method and implemented on FPGA. In comparison to the Sobel algorithm, the proposed method has superior anti-noise capability.
This paper presents a new image description and matching process based on internal self-similarity property of images. Various definitions of self-similarity are explored to find the best one for image matching. The m...
详细信息
image registration has been a broadly applied topic across the photogrammetric/remote sensing and computer vision communities. It is a foundational step for many applications such geopositioning, data fusion, change d...
详细信息
The image classification process is based on the assumption that pixels which have similar spatial distribution patterns, or statistical characteristics, belong to the same spectral class. In a previous study we have ...
详细信息
ISBN:
(数字)9783642215933
ISBN:
(纸本)9783642215926;9783642215933
The image classification process is based on the assumption that pixels which have similar spatial distribution patterns, or statistical characteristics, belong to the same spectral class. In a previous study we have shown how we can improve the accuracy of classification of remotely sensed imagery data by incorporating contextual elevation knowledge in a form of a digital elevation model with the output of the classification process using Dempster-Shafer Theory of Evidence. A knowledge based approach is created for this purpose using suitable production rules derived from the elevation distributions and range of values for the elevation data attached to a particular satellite image. Production rules are the major part of knowledge representation and have the basic form: IF condition THEN Inference. Although the basic form of production rules has shown accuracy improvement, in general, in some cases accuracy can degrade. In this paper we propose a "refined" approach that takes into account the actual "distribution" of elevation values for each class rather than simply the "range" of values to solve the accuracy degradation. This approach is performed by refining the basic production rules used in the previous study taking into account the number of pixels at each elevation within the elevation distribution for each class.
暂无评论