检索结果-内蒙古大学图书馆

Elastic temporal alignment for few-shot action recognition

IET computer vision 2023年第1期17卷 39-50页

作者： Pan, Fei Xu, Chunlei Zhang, Hongjie Guo, Jie Guo, Yanwen Nanjing Univ Natl Key Lab Novel Software Technol 163 Xianlin Ave Nanjing Peoples R China

Few-shot action recognition aims to learn a classification model with good generalisation ability when trained with only a few labelled videos. However, it is difficult to learn discriminative feature representations for videos in such a setting. The Elastic Temporal Alignment (ETA) for few-shot action recognition is proposed. First, a convolutional neural network is employed to extract feature representations of video frames sparsely sampled from videos. In order to obtain the similarity of two videos, a temporal alignment estimation function is utilised to estimate the matching score between each pair of frames from the two videos through an elastic alignment mechanism. The analysis shows that when we judge whether two frames from respective videos are matched, multiple adjacent frames in the videos should be considered, so as to embody the temporal information. Thus, before feeding per-frame feature vectors of videos into the temporal alignment estimation function, a temporal message passing function is leveraged to propagate the information of per-frame features in the temporal domain. The method has been evaluated on four action recognition datasets, including Kinetics, Something-Something V2, HMDB51, and UCF101. The experimental results verify the effectiveness of ETA and show its superiority over state-of-the-art methods.

关键词： temporal alignment estimation function feature extraction temporal message passing function video signal processing image representation learning (artificial intelligence) Neural nets elastic alignment mechanism per-frame feature vectors message passing multiple adjacent frames action recognition datasets computer vision and image processing techniques Elastic Temporal Alignment convolutional neural nets per-frame features Elastic temporal alignment respective videos image recognition temporal information Video signal processing temporal domain labelled videos few-shot action recognition aims discriminative feature representations image motion analysis video frames

来源：评论

学校读者我要写书评

暂无评论

Texture classification using invariant features of local textures

引用

IET image processing 2010年第3期4卷 158-171页

作者： Janney, P. Geers, G. Univ New S Wales Sch Engn & Comp Sci Sydney NSW 2032 Australia NICTA Sydney NSW 2032 Australia

In this paper, the authors present a texture descriptor algorithm called invariant features of local textures (IFLT). IFLT generates scale, rotation and (essentially) illumination invariant descriptors from a small neighbourhood of pixels around a centre pixel or a texture patch. Texture classification experiments were carried out on the Brodatz, Outex and KTH-TIPS2 databases. Demonstrated texture classification accuracy exceeds the previously published state of the art at a significantly lower computational cost. Experiments also suggests that IFLT descriptors are in a sense intuitive texture descriptors.

关键词： image resolution texture patch computer vision and image processing techniques image classification invariant features of local textures feature extraction texture descriptor algorithm texture classification image texture Optical, image and video signal processing illumination invariant descriptors KTH-TIPS2 databases

来源：评论

学校读者我要写书评

暂无评论

Salient object detection based on edge-interior feature fusion

引用

IET image processing 2023年第2期17卷 337-348页

作者： Shi, Yadi Qin, Guihe Liang, Yanhua Wang, Xinchao Yan, Jie Zhang, Zhonghan Jilin Univ Coll Software Changchun Peoples R China Jilin Univ Coll Comp Sci & Technol Changchun 130012 Peoples R China Jilin Univ Minist Educ Key Lab Symbol Computat & Knowledge Engn Changchun Peoples R China

Recently, existing FCNs-based methods have shown their advantages in processing object boundaries. However, these methods still suffer from false object interference, which appears in saliency predictions. To solve this problem, an edge-interior feature fusion (EIFF) framework is proposed, which consists of an internal-boundary decoupled generation structure with receptive field enlargement and attention mechanism enhancement, and a salient feature refinement module. Specifically, the framework first learns edge features and interior features through an internal-boundary decoupling generation network, which is supervised by labels obtained by decoupling ground-truth through an image erosion algorithm. Then, feature refinement module (FRM) is designed to purify the coarse prediction by focusing on the ambiguous regions through a mining strategy to generate the final saliency map. To compensate for shortcomings of the BCE and IU loss, we also introduce a weighted loss to guide our model to focus more on the error-prone parts. Experimental results on five benchmark datasets demonstrate that the proposed method performs favorably against 19 state-of-the-art approaches under four standard metrics.

关键词： image segmentation feature extraction salient object detection learning (artificial intelligence) edge-interior feature fusion framework Machine learning (artificial intelligence) FCN-based methods saliency predictions EIFF image fusion computer vision and image processing techniques FRM interior features BCE data mining coarse prediction internal-boundary decoupled generation structure internal-boundary decoupling generation network object detection image recognition salient feature refinement module Data mining false object interference image erosion algorithm edge detection receptive field enlargement IU loss attention mechanism enhancement object boundaries

来源：评论

学校读者我要写书评

暂无评论

Measuring empirical discrepancy in image segmentation results

引用

IET computer vision 2012年第3期6卷 224-230页

作者： Correa-Tome, F. E. Sanchez-Yanez, R. E. Ayala-Ramirez, V. Univ Guanajuato DICIS Salamanca Gto 36885 Mexico

A methodology for comparison of boundary and segmentation images based on Precision-Recall graphs is presented in this study. The proposed methodology compares the location of edge pixels between an image under test and an ideal reference, in order to obtain a precise normalised similarity measure. This approach also deals with the case when multiple references are available using a merging procedure. Small displacement errors in edge pixel location are handled using a tolerance radius, which introduces the problem of multiple matching between test and reference edge pixels. This problem is addressed as a bipartite graph, solved by using the Hopcroft-Karp algorithm to obtain the maximum number of unique matchings. Experiments have been carried out in order to determine the performance of this evaluation approach.

关键词： image segmentation graph theory edge pixels precision-recall graphs bipartite graph image matching displacement errors empirical discrepancy measurement Hopcroft-Karp algorithm Combinatorial mathematics Optical, image and video signal processing computer vision and image processing techniques merging procedure

来源：评论

学校读者我要写书评

暂无评论

Optimal look-up table-based data hiding

引用

IET SIGNAL processing 2011年第2期5卷 171-179页

作者： Wang, X. Zhang, X. -P. Ryerson Univ Dept Elect & Comp Engn Toronto ON M5B 2K3 Canada

In this study, the authors present a novel data hiding scheme using the minimum distortion look-up table (LUT) embedding that achieves good distortion-robustness performance. LUT-based data hiding is a simple and efficient way to embed information into multimedia content for various applications, such as transaction tracking and database annotation. The authors find it possible to optimally reduce the data hiding-introduced distortion by designing the LUT according to the distribution of the host at a given robustness level. The authors first analyse the distortion introduced by LUT embedding and formulate its relationship with run constraints of LUT to construct an optimal coding problem. Subsequently, a Viterbi algorithm is presented to find the minimum distortion LUT. Then a new practical data hiding scheme using the optimal LUT is applied in the wavelet domain. Theoretical analysis and numerical results show that the new LUT design achieves not only less distortion but also more robustness than the traditional LUT-based data embedding schemes under common attacks such as Gaussian noise and JPEG compression.

关键词： computer vision and image processing techniques data compression image and video coding wavelet transforms telecommunication security table lookup Codes image coding distortion Viterbi coding Viterbi decoding JPEG compression LUT-based data hiding database annotation wavelet domain look-up table Integral transforms Data handling techniques transaction tracking

来源：评论

学校读者我要写书评

暂无评论

image denoising method based on variable exponential fractional-integer-order total variation and tight frame sparse regularization

引用

IET image processing 2021年第1期15卷 101-114页

作者： Wang, Yingmei Wang, Zhendong Shandong Univ Technol Sch Math & Stat Zibo 255049 Peoples R China Shandong Univ Sch Math Jinan Peoples R China

This study presents a variational image restoration algorithm based on variable exponential fractional-order total variation (TV), variable exponential integer-order TV and tight frame sparse regularization. The energy functional of this variational problem is composed of four parts: a fractional-order TV regularization term with a variable exponent, an integer-order TV regularization term with a variable exponent, a data fidelity term and a tight frame regularization term. The variable exponent is a function of the gradient information of the image. The first three parts of the energy functional is the smooth part, and the last part is the non-smooth part. To minimize the smooth part, the optimization problem can be simply transformed into a gradient descent flow using the variational method. For the non-smooth part, we use the soft-thresholding method over the sparse framelet coefficients. As the combination of the fractional-order derivative and integer-order derivative, the variable exponent and the sparse regularisation, the proposed method can effectively remove noise of the image, protect the image boundary, and keep image texture details well. Experiments with simulated data and real data are provided to validate the effectiveness of proposed method. This method is robust to noise and is of some practical application value.

关键词： variational techniques integer-order TV regularization term image denoising variable exponent computer vision and image processing techniques variable exponential fractional-integer-order total variation tight frame sparse regularization image restoration fractional-order derivative gradient descent flow energy functional fractional-order TV regularization term image texture variational image restoration integer-order derivative Optimisation techniques gradient methods Optical, image and video signal processing

来源：评论

学校读者我要写书评

暂无评论

Matching parameter estimation by using the radial basis function interpolation

引用

IET image processing 2012年第4期6卷 407-416页

作者： Fuji, Y. Abe, Y. Iiguni, Y. Osaka Univ Grad Sch Engn Sci Toyonaka Osaka 5608531 Japan

A matching parameter estimation method with subpixel accuracy is derived by using the radial basis function (RBF) interpolation. This method reconstructs two analogue images from two given digital images by the RBF, and then minimises a non-linear cost function by the steepest-descent algorithm to estimate translation, rotation, scaling factor and intensity change between the two analogue images. The RBF provides accurate interpolation, resulting in accurate estimation. A Gaussian weighting function is introduced into the cost function to provide a local estimate within a region of interest (ROC). Then double integrals included in the cost function are analytically computed and the computational complexity is significantly reduced by exploiting the property that the Gaussian function decays rapidly. When the matching parameters are not constant over the whole image, or equivalently, the ROC is set to be small, the proposed method is better than the conventional phase correlation method in estimation accuracy.

关键词： Gaussian weighting function rotation estimation matching parameter estimation method image registration double-integrals image reconstruction scaling factor nonlinear cost function radial basis function interpolation computer vision and image processing techniques intensity change translation estimation Optical, image and video signal processing phase correlation method analogue image reconstruction image matching subpixel accuracy digital images Interpolation and function approximation (numerical analysis) RBF interpolation interpolation steepest-descent algorithm computational complexity estimation accuracy gradient methods

来源：评论

学校读者我要写书评

暂无评论

Full-scale attention network for automated organ segmentation on head and neck CT and MR images

引用

IET image processing 2023年第3期17卷 660-673页

作者： Zhong, Zhiqiang He, Lian Chen, Changxiu Yang, Xingli Lin, Li Yan, Ziye Tian, Mengqiu Sun, Ying Zhan, Yinwei Guangdong Univ Technol Sch Comp Sci Guangzhou 510006 Peoples R China Guangzhou Percept Vis Med Technol Guangzhou 510275 Peoples R China Sun Yat Sen Univ Dept Radiat Oncol Affiliated Hosp 1 Guangzhou Peoples R China Sun Yat Sen Univ Guangdong Key Lab Nasopharyngeal Carcinoma Diag & Collaborat Innovat Ctr Canc MedCanc Ctr State Key Lab Oncol South ChinaDept Radiat Oncol Guangzhou 510060 Peoples R China

MRI and CT images have been routinely used in clinical practice for treatment planning of the head-and-neck (HAN) radiotherapy. Delineating organs-at-risk (OAR) is an essential step in radiotherapy, however, it is time-consuming and prone to inter-observer variation. The existing automatic segmentation approaches are either limited by image registration or lack of global spatial awareness, thus under-performed when dealing with segmentation of complex anatomies. Herein, we propose a full-scale attention network (FSANet) that integrates bi-side skip connections, full-scale feature fusion modules (FFM), a feature pyramid fusion and supervision module (FPFSM) to accurately and efficiently delineate OARs in HAN region on CT and MRI scans. Specifically, bi-side skip connections were adopted to keep small targets in the deep network and to capture semantic features at different scales. The FFM with cascaded attention mechanisms were used to recalibrate the significant channels and salient regions in the feature maps. The FPFSM was used to guide the network to learn the hierarchical representation so as to improve the segmentation robustness. The proposed algorithm was validated on the public benchmark HAN CT dataset and an in-house MR dataset. Both results show significant improvement compared to state-of-the-art OAR single-stage segmentation methods for the HAN region.

关键词： efficiently delineate OARs Biology and medical computing full-scale attention network image segmentation feature extraction feature maps biomedical MRI Medical magnetic resonance imaging and spectroscopy deep network FFM X-rays and particle beams (medical uses) global spatial awareness Neural nets Optical, image and video signal processing image fusion FPFSM computer vision and image processing techniques clinical practice Biomedical magnetic resonance imaging and spectroscopy treatment planning segmentation robustness inter-observer variation X-ray techniques: radiography and computed tomography (biomedical imaging/measurement) full-scale feature fusion modules HAN region semantic features feature pyramid fusion image registration radiation therapy supervision module automated organ segmentation existing automatic segmentation approaches computerised tomography organs-at-risk state-of-the-art OAR single-stage segmentation methods bi-side skip connections biological organs radiotherapy cascaded attention mechanisms Patient diagnostic methods and instrumentation public benchmark HAN CT dataset Radiation therapy medical image processing

来源：评论

学校读者我要写书评

暂无评论

AFF-UNIT: Adaptive feature fusion for unsupervised image-to-image translation

引用

IET image processing 2021年第13期15卷 3172-3188页

作者： Li, Yuqiang Meng, Haochen Lin, Hong Liu, Chun Wuhan Univ Technol Sch Comp Sci & Technol Wuhan Peoples R China

The task of image-to-image translation is to generate images closer to the target domain style while preserving the significant features of the original image. This paper contends an adaptive feature fusion method for unsupervised image translation. The proposed architecture, termed as AFF-UNIT, is based on a compact network structure to further improve the quality of generated images. First of all, a feature extraction module based on an adaptive feature fusion method is proposed, which combines low-level fine-grained information and high-level semantic information to obtain feature maps with richer information. At the same time, a feature-similarity loss is proposed to guide the feature extraction module to extract features that are more conducive to improving the translation result. In addition, AFF-UNIT reuses the feature extraction module in the generator and discriminator to simplify the framework. Extensive experiments on five popular benchmarks demonstrate the superior performance of AFF-UNIT over state-of-the-art methods in terms of FID, KID, IS, and also human preference. Comprehensive ablation studies are also carried out to isolate the validity of each proposed component.

关键词： Optical, image and video signal processing Sensor fusion computer vision and image processing techniques

来源：评论

学校读者我要写书评

暂无评论

Vehicle detection in aerial images based on lightweight deep convolutional network

引用

IET image processing 2021年第2期15卷 479-491页

作者： Shen, Jiaquan Liu, Ningzhong Sun, Han Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing Peoples R China MIIT Key Lab Pattern Anal & Machine Intelligence Nanjing Peoples R China

Vehicle detection in aerial images is an interesting and challenging task. Traditional methods are based on sliding-window search and handcrafted features, which limits the representation power and has heavy computational costs. Recent research have shown that deep-learning algorithms are widely used in the field of object detection. However, the deep-learning algorithms still face many difficulties and challenges in the object detection under the aerial scene. Meanwhile, the high computational cost of detection algorithms lead to low-detection efficiency. In this study, we build a fast and accurate lightweight detection framework for vehicle detection in aerial scenes. The proposed detection method improves the expressive ability of detection network and significantly reduces the amount of calculations in the model. Meanwhile, setting suitable anchor boxes according to the size of the object vehicles have been introduced in our model, which also effectively improves the performance of the detection. In addition, we have published a new aerial vehicle image dataset and verified the effectiveness of our method. In the Munich dataset and our dataset, our method achieves 85.8% and 91.2% of the mean average precision (mAP), and its detection time is 1.78 and 0.048 s on Nvidia Titan XP. Our results show that the proposed framework achieves significant improvement over several alternatives and state-of-the-art schemes with higher accuracy and less detection time.

关键词： Optical, image and video signal processing computer vision and image processing techniques Aerospace engineering computing Neural nets

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：