Recent deep neural network(DNN)based blind image quality assessment(BIQA)approaches take mean opinion score(MOS)as ground-truth labels,which would lead to cross-datasets biases and limited generalization ability of th...Recent deep neural network(DNN)based blind image quality assessment(BIQA)approaches take mean opinion score(MOS)as ground-truth labels,which would lead to cross-datasets biases and limited generalization ability of the DNN-based BIQA model.This work validates the natural instability of MOS through investigating the neuropsychological characteristics inside the human visual system during quality perception.By combining persistent homology analysis with electroencephalogram(EEG),the physiologically meaningful features of the brain responses to different distortion levels are extracted.The physiological features indicate that although volunteers view exactly the same image content,their EEG features are quite varied.Based on the physiological results,we advocate treating MOS as noisy labels and optimizing the DNN based BIQA model with earlystop strategies.Experimental results on both innerdataset and cross-dataset demonstrate the superiority of our optimization approach in terms of generalization ability.展开更多
Blind image quality assessment(BIQA)is of fundamental importance in low-level computer vision community.Increasing interest has been drawn in exploiting deep neural networks for BIQA.Despite of the notable success ach...Blind image quality assessment(BIQA)is of fundamental importance in low-level computer vision community.Increasing interest has been drawn in exploiting deep neural networks for BIQA.Despite of the notable success achieved,there is a broad consensus that training deep convolutional neural networks(DCNN)heavily relies on massive annotated data.Unfortunately,BIQA is typically a small sample problem,resulting the generalization ability of BIQA severely restricted.In order to improve the accuracy and generalization ability of BIQA metrics,this work proposed a totally opinion-unaware BIQA in which no subjective annotations are involved in the training stage.Multiple full-reference image quality assessment(FR-IQA)metrics are employed to label the distorted image as a substitution of subjective quality annotation.A deep neural network(DNN)is trained to blindly predict the multiple FR-IQA score in absence of corresponding pristine image.In the end,a selfsupervised FR-IQA score aggregator implemented by adversarial auto-encoder pools the predictions of multiple FR-IQA scores into the final quality predicting score.Even though none of subjective scores are involved in the training stage,experimental results indicate that our proposed full reference induced BIQA framework is as competitive as state-of-the-art BIQA metrics.展开更多
Virtual reality(VR) environment can provide immersive experience to viewers.Under the VR environment, providing a good quality of experience is extremely important.Therefore, in this paper, we present an image quality...Virtual reality(VR) environment can provide immersive experience to viewers.Under the VR environment, providing a good quality of experience is extremely important.Therefore, in this paper, we present an image quality assessment(IQA) study on omnidirectional images. We first build an omnidirectional IQA(OIQA) database, including 16 source images with their corresponding 320 distorted images. We add four commonly encountered distortions. These distortions are JPEG compression, JPEG2000 compression, Gaussian blur, and Gaussian noise. Then we conduct a subjective quality evaluation study in the VR environment based on the OIQA database. Considering that visual attention is more important in VR environment, head and eye movement data are also tracked and collected during the quality rating experiments. The 16 raw and their corresponding distorted images,subjective quality assessment scores, and the head-orientation data and eye-gaze data together constitute the OIQA database. Based on the OIQA database, we test some state-of-the-art full-reference IQA(FR-IQA) measures on equirectangular format or cubic formatomnidirectional images. The results show that applying FR-IQA metrics on cubic format omnidirectional images could improve their performance. The performance of some FR-IQA metrics combining the saliency weight of three different types are also tested based on our database. Some new phenomena different from traditional IQA are observed.展开更多
Image quality assessment(IQA)is constantly innovating,but there are still three types of stickers that have not been resolved:the“content sticker”-limitation of training set,the“annotation sticker”-subjective inst...Image quality assessment(IQA)is constantly innovating,but there are still three types of stickers that have not been resolved:the“content sticker”-limitation of training set,the“annotation sticker”-subjective instability in opinion scores and the“distortion sticker”-disordered distortion settings.In this paper,a No-Reference Image Quality Assessment(NR IQA)approach is proposed to deal with the problems.For“content sticker”,we introduce the idea of pairwise comparison and generate a largescale ranking set to pre-train the network;For“annotation sticker”,the absolute noise-containing subjective scores are transformed into ranking comparison results,and we design an indirect unsupervised regression based on EigenValue Decomposition(EVD);For“distortion sticker”,we propose a perception-based distortion classification method,which makes the distortion types clear and refined.Experiments have proved that our NR IQA approach Experiments show that the algorithm performs well and has good generalization ability.Furthermore,the proposed perception based distortion classification method would be able to provide insights on how the visual related studies may be developed and to broaden our understanding of human visual system.展开更多
Recent studies on no-reference image quality assessment (NR-IQA) methods usually learn to evaluate the image quality by regressing from human subjective scores of the training samples. This study presented an NR-IQA m...Recent studies on no-reference image quality assessment (NR-IQA) methods usually learn to evaluate the image quality by regressing from human subjective scores of the training samples. This study presented an NR-IQA method based on the basic image visual parameters without using human scored image databases in learning. We demonstrated that these features comprised the most basic characteristics for constructing an image and influencing the visual quality of an image. In this paper, the definitions, computational method, and relationships among these visual metrics were described. We subsequently proposed a no-reference assessment function, which was referred to as a visual parameter measurement index (VPMI), based on the integration of these visual metrics to assess image quality. It is established that the maximum of VPMI corresponds to the best quality of the color image. We verified this method using the popular assessment database—image quality assessment database (LIVE), and the results indicated that the proposed method matched better with the subjective assessment of human vision. Compared with other image quality assessment models, it is highly competitive. VPMI has low computational complexity, which makes it promising to implement in real-time image assessment systems.展开更多
A new method for no-reference image quality assessment based on hybrid fuzzy-genetic technique is proposed. Noise variance and edge sharpness level of the restored image are two basic metrics for assessing the perform...A new method for no-reference image quality assessment based on hybrid fuzzy-genetic technique is proposed. Noise variance and edge sharpness level of the restored image are two basic metrics for assessing the performance of the restoration algorithm, then a fuzzy if-then inference system is developed to combine the two metrics to get a final quality score, and the parameters of the fuzzy membership function are trained with genetic algorithms. Experiments results show that the image quality score correlates well with mean opinion score and the proposed approach is robust and effective.展开更多
Image enhancement is a popular technique,which is widely used to improve the visual quality of images.While image enhancement has been extensively investigated,the relevant quality assessment of enhanced images remain...Image enhancement is a popular technique,which is widely used to improve the visual quality of images.While image enhancement has been extensively investigated,the relevant quality assessment of enhanced images remains an open problem,which may hinder further development of enhancement techniques.In this paper,a no-reference quality metric for digitally enhanced images is proposed.Three kinds of features are extracted for characterizing the quality of enhanced images,including non-structural information,sharpness and naturalness.Specifically,a total of 42 perceptual features are extracted and used to train a support vector regression(SVR) model.Finally,the trained SVR model is used for predicting the quality of enhanced images.The performance of the proposed method is evaluated on several enhancement-related databases,including a new enhanced image database built by the authors.The experimental results demonstrate the efficiency and advantage of the proposed metric.展开更多
Considering that there is no single full reference image quality assessment method that could give the best performance in all situations, some multi-method fusion metrics were proposed. Machine learning techniques ar...Considering that there is no single full reference image quality assessment method that could give the best performance in all situations, some multi-method fusion metrics were proposed. Machine learning techniques are often involved in such multi-method fusion metrics so that its output would be more consistent with human visual perceptions. On the other hand, the robustness and generalization ability of these multi-method fusion metrics are questioned because of the scarce of images with mean opinion scores. In order to comprehensively validate whether or not the generalization ability of such multi-method fusion IQA metrics are satisfying, we construct a new image database which contains up to 60 reference images. The newly built image database is then used to test the generalization ability of different multi-method fusion IQA metrics. Cross database validation experiment indicates that in our new image database, the performances of all the multi-method fusion IQA metrics have no statistical significant different with some single-method IQA metrics such as FSIM and MAD. In the end, a thorough analysis is given to explain why the performance of multi-method fusion IQA framework drop significantly in cross database validation.展开更多
This paper presents an investigation on the effect of JPEG compression on the similarity between the target image and the background,where the similarity is further used to determine the degree of clutter in the image...This paper presents an investigation on the effect of JPEG compression on the similarity between the target image and the background,where the similarity is further used to determine the degree of clutter in the image.Four new clutter metrics based on image quality assessment are introduced,among which the Haar wavelet-based perceptual similarity index,known as HaarPSI,provides the best target acquisition prediction results.It is shown that the similarity between the target and the background at the boundary between visually lossless and visually lossy compression does not change significantly compared to the case when an uncompressed image is used.In future work,through subjective tests,it is necessary to check whether this presence of compression at the threshold of just noticeable differences will affect the human target acquisition performance.Similarity values are compared with the results of subjective tests of the well-known target Search_2 database,where the degree of agreement between objective and subjective scores,measured through linear correlation,reached a value of 90%.展开更多
While quality assessment is essential for testing, optimizing, benchmarking, monitoring, and inspecting related systems and services, it also plays an essential role in the design of virtually all visual signal proces...While quality assessment is essential for testing, optimizing, benchmarking, monitoring, and inspecting related systems and services, it also plays an essential role in the design of virtually all visual signal processing and communication algorithms, as well as various related decision-making processes. In this paper, we first provide an overview of recently derived quality assessment approaches for traditional visual signals (i.e., 2D images/videos), with highlights for new trends (such as machine learning approaches). On the other hand, with the ongoing development of devices and multimedia services, newly emerged visual signals (e.g., mobile/3D videos) are becoming more and more popular. This work focuses on recent progresses of quality metrics, which have been reviewed for the newly emerged forms of visual signals, which include scalable and mobile videos, High Dynamic Range (HDR) images, image segmentation results, 3D images/videos, and retargeted images.展开更多
The bandwidth of internet connections is still a bottleneck when transmitting large amounts of images,making the image quality assessment essential.Neurophysiological assessment of image quality has highlight advantag...The bandwidth of internet connections is still a bottleneck when transmitting large amounts of images,making the image quality assessment essential.Neurophysiological assessment of image quality has highlight advantages for it does not interfere with natural viewing behavior.However,in JPEG compression,the previous study is hard to tell the difference between the electroencephalogram(EEG)evoked by different quality images.In this paper,we propose an EEG analysis approach based on algebraic topology analysis,and the result shows that the difference between Euler characteristics of EEG evoked by different distortion images is striking both in the alpha and beta band.Moreover,we further discuss the relationship between the images and the EEG signals,and the results implied that the algebraic topological properties of images are consistent with that of brain perception,which is possible to give birth to braininspired image compression based on algebraic topological features.In general,an algebraic topologybased approach was proposed in this paper to analyze the perceptual characteristics of image quality,which will be beneficial to provide a reliable score for data compression in the network and improve the network transmission capacity.展开更多
In this paper, we propose a novel image recompression frame- work and image quality assessment （IQA） method to efficiently recompress Internet images. With this framework image size is significantly reduced without ...In this paper, we propose a novel image recompression frame- work and image quality assessment （IQA） method to efficiently recompress Internet images. With this framework image size is significantly reduced without affecting spatial resolution or perceptible quality of the image. With the help of IQA, the relationship between image quality and image evaluation scores can be quickly established, and the optimal quality factor can be obtained quickly and accurately within a pre - determined perceptual quality range. This process ensures the image＇s perceptual quality, which is applied to each input image. The test results show that, using the proposed method, the file size of images can be reduced by about 45%-60% without affecting their visual quality. Moreover, our new image -reeompression framework can be used in to many different application scenarios.展开更多
本文提出了一种基于双交叉注意力融合的Swin-AK Transformer(Swin Transformer based on alterable kernel convolution)和手工特征相结合的智能手机拍摄图像质量评价方法。首先,提取了影响图像质量的手工特征,这些特征可以捕捉到图像...本文提出了一种基于双交叉注意力融合的Swin-AK Transformer(Swin Transformer based on alterable kernel convolution)和手工特征相结合的智能手机拍摄图像质量评价方法。首先,提取了影响图像质量的手工特征,这些特征可以捕捉到图像中细微的视觉变化;其次,提出了Swin-AK Transformer,增强了模型对局部信息的提取和处理能力。此外,本文设计了双交叉注意力融合模块,结合空间注意力和通道注意力机制,融合了手工特征与深度特征,实现了更加精确的图像质量预测。实验结果表明,在SPAQ和LIVE-C数据集上,皮尔森线性相关系数分别达到0.932和0.885,斯皮尔曼等级排序相关系数分别达到0.929和0.858。上述结果证明了本文提出的方法能够有效地预测智能手机拍摄图像的质量。展开更多
Neural network methods have recently emerged as a hot topic in computed tomography(CT) imaging owing to their powerful fitting ability;however, their potential applications still need to be carefully studied because t...Neural network methods have recently emerged as a hot topic in computed tomography(CT) imaging owing to their powerful fitting ability;however, their potential applications still need to be carefully studied because their results are often difficult to interpret and are ambiguous in generalizability. Thus, quality assessments of the results obtained from a neural network are necessary to evaluate the neural network. Assessing the image quality of neural networks using traditional objective measurements is not appropriate because neural networks are nonstationary and nonlinear. In contrast, subjective assessments are trustworthy, although they are time-and energy-consuming for radiologists. Model observers that mimic subjective assessment require the mean and covariance of images, which are calculated from numerous image samples;however, this has not yet been applied to the evaluation of neural networks. In this study, we propose an analytical method for noise propagation from a single projection to efficiently evaluate convolutional neural networks(CNNs) in the CT imaging field. We propagate noise through nonlinear layers in a CNN using the Taylor expansion. Nesting of the linear and nonlinear layer noise propagation constitutes the covariance estimation of the CNN. A commonly used U-net structure is adopted for validation. The results reveal that the covariance estimation obtained from the proposed analytical method agrees well with that obtained from the image samples for different phantoms, noise levels, and activation functions, demonstrating that propagating noise from only a single projection is feasible for CNN methods in CT reconstruction. In addition, we use covariance estimation to provide three measurements for the qualitative and quantitative performance evaluation of U-net. The results indicate that the network cannot be applied to projections with high noise levels and possesses limitations in terms of efficiency for processing low-noise projections. U-net is more effective in improving the image quality of smooth regions compared with that of the edge. LeakyReLU outperforms Swish in terms of noise reduction.展开更多
基金supported by the Medium and Long-term Science and Technology Plan for Radio,Television,and Online Audiovisuals(2023AC0200)the Public Welfare Technology Application Research Project of Zhejiang Province,China(No.LGF21F010001).
文摘Recent deep neural network(DNN)based blind image quality assessment(BIQA)approaches take mean opinion score(MOS)as ground-truth labels,which would lead to cross-datasets biases and limited generalization ability of the DNN-based BIQA model.This work validates the natural instability of MOS through investigating the neuropsychological characteristics inside the human visual system during quality perception.By combining persistent homology analysis with electroencephalogram(EEG),the physiologically meaningful features of the brain responses to different distortion levels are extracted.The physiological features indicate that although volunteers view exactly the same image content,their EEG features are quite varied.Based on the physiological results,we advocate treating MOS as noisy labels and optimizing the DNN based BIQA model with earlystop strategies.Experimental results on both innerdataset and cross-dataset demonstrate the superiority of our optimization approach in terms of generalization ability.
基金supported by the Public Welfare Technology Application Research Project of Zhejiang Province,China(No.LGF21F010001)the Key Research and Development Program of Zhejiang Province,China(Grant No.2019C01002)the Key Research and Development Program of Zhejiang Province,China(Grant No.2021C03138)。
文摘Blind image quality assessment(BIQA)is of fundamental importance in low-level computer vision community.Increasing interest has been drawn in exploiting deep neural networks for BIQA.Despite of the notable success achieved,there is a broad consensus that training deep convolutional neural networks(DCNN)heavily relies on massive annotated data.Unfortunately,BIQA is typically a small sample problem,resulting the generalization ability of BIQA severely restricted.In order to improve the accuracy and generalization ability of BIQA metrics,this work proposed a totally opinion-unaware BIQA in which no subjective annotations are involved in the training stage.Multiple full-reference image quality assessment(FR-IQA)metrics are employed to label the distorted image as a substitution of subjective quality annotation.A deep neural network(DNN)is trained to blindly predict the multiple FR-IQA score in absence of corresponding pristine image.In the end,a selfsupervised FR-IQA score aggregator implemented by adversarial auto-encoder pools the predictions of multiple FR-IQA scores into the final quality predicting score.Even though none of subjective scores are involved in the training stage,experimental results indicate that our proposed full reference induced BIQA framework is as competitive as state-of-the-art BIQA metrics.
文摘Virtual reality(VR) environment can provide immersive experience to viewers.Under the VR environment, providing a good quality of experience is extremely important.Therefore, in this paper, we present an image quality assessment(IQA) study on omnidirectional images. We first build an omnidirectional IQA(OIQA) database, including 16 source images with their corresponding 320 distorted images. We add four commonly encountered distortions. These distortions are JPEG compression, JPEG2000 compression, Gaussian blur, and Gaussian noise. Then we conduct a subjective quality evaluation study in the VR environment based on the OIQA database. Considering that visual attention is more important in VR environment, head and eye movement data are also tracked and collected during the quality rating experiments. The 16 raw and their corresponding distorted images,subjective quality assessment scores, and the head-orientation data and eye-gaze data together constitute the OIQA database. Based on the OIQA database, we test some state-of-the-art full-reference IQA(FR-IQA) measures on equirectangular format or cubic formatomnidirectional images. The results show that applying FR-IQA metrics on cubic format omnidirectional images could improve their performance. The performance of some FR-IQA metrics combining the saliency weight of three different types are also tested based on our database. Some new phenomena different from traditional IQA are observed.
基金supported by the Specialized Research Fund for the Doctoral Program of Higher Education of China, "Research of Visual Perception for Impairments of Color Information in High-Definition Images" (No.20110018110001)
文摘Image quality assessment(IQA)is constantly innovating,but there are still three types of stickers that have not been resolved:the“content sticker”-limitation of training set,the“annotation sticker”-subjective instability in opinion scores and the“distortion sticker”-disordered distortion settings.In this paper,a No-Reference Image Quality Assessment(NR IQA)approach is proposed to deal with the problems.For“content sticker”,we introduce the idea of pairwise comparison and generate a largescale ranking set to pre-train the network;For“annotation sticker”,the absolute noise-containing subjective scores are transformed into ranking comparison results,and we design an indirect unsupervised regression based on EigenValue Decomposition(EVD);For“distortion sticker”,we propose a perception-based distortion classification method,which makes the distortion types clear and refined.Experiments have proved that our NR IQA approach Experiments show that the algorithm performs well and has good generalization ability.Furthermore,the proposed perception based distortion classification method would be able to provide insights on how the visual related studies may be developed and to broaden our understanding of human visual system.
基金supported by the National Natural Science Foundation of China under Grants No.61773094,No.61573080,No.91420105,and No.61375115National Program on Key Basic Research Project(973 Program)under Grant No.2013CB329401+1 种基金National High-Tech R&D Program of China(863 Program)under Grant No.2015AA020505Sichuan Province Science and Technology Project under Grants No.2015SZ0141 and No.2018ZA0138
文摘Recent studies on no-reference image quality assessment (NR-IQA) methods usually learn to evaluate the image quality by regressing from human subjective scores of the training samples. This study presented an NR-IQA method based on the basic image visual parameters without using human scored image databases in learning. We demonstrated that these features comprised the most basic characteristics for constructing an image and influencing the visual quality of an image. In this paper, the definitions, computational method, and relationships among these visual metrics were described. We subsequently proposed a no-reference assessment function, which was referred to as a visual parameter measurement index (VPMI), based on the integration of these visual metrics to assess image quality. It is established that the maximum of VPMI corresponds to the best quality of the color image. We verified this method using the popular assessment database—image quality assessment database (LIVE), and the results indicated that the proposed method matched better with the subjective assessment of human vision. Compared with other image quality assessment models, it is highly competitive. VPMI has low computational complexity, which makes it promising to implement in real-time image assessment systems.
文摘A new method for no-reference image quality assessment based on hybrid fuzzy-genetic technique is proposed. Noise variance and edge sharpness level of the restored image are two basic metrics for assessing the performance of the restoration algorithm, then a fuzzy if-then inference system is developed to combine the two metrics to get a final quality score, and the parameters of the fuzzy membership function are trained with genetic algorithms. Experiments results show that the image quality score correlates well with mean opinion score and the proposed approach is robust and effective.
基金supported in part by the National Natural Science Foundation of China under Grant 61379143in part by the Fundamental Research Funds for the Central Universities under Grant 2015QNA66in part by the Qing Lan Project of Jiangsu Province
文摘Image enhancement is a popular technique,which is widely used to improve the visual quality of images.While image enhancement has been extensively investigated,the relevant quality assessment of enhanced images remains an open problem,which may hinder further development of enhancement techniques.In this paper,a no-reference quality metric for digitally enhanced images is proposed.Three kinds of features are extracted for characterizing the quality of enhanced images,including non-structural information,sharpness and naturalness.Specifically,a total of 42 perceptual features are extracted and used to train a support vector regression(SVR) model.Finally,the trained SVR model is used for predicting the quality of enhanced images.The performance of the proposed method is evaluated on several enhancement-related databases,including a new enhanced image database built by the authors.The experimental results demonstrate the efficiency and advantage of the proposed metric.
基金supported by “the Fundamental Research Funds for the Central Universities” No.2018CUCTJ081
文摘Considering that there is no single full reference image quality assessment method that could give the best performance in all situations, some multi-method fusion metrics were proposed. Machine learning techniques are often involved in such multi-method fusion metrics so that its output would be more consistent with human visual perceptions. On the other hand, the robustness and generalization ability of these multi-method fusion metrics are questioned because of the scarce of images with mean opinion scores. In order to comprehensively validate whether or not the generalization ability of such multi-method fusion IQA metrics are satisfying, we construct a new image database which contains up to 60 reference images. The newly built image database is then used to test the generalization ability of different multi-method fusion IQA metrics. Cross database validation experiment indicates that in our new image database, the performances of all the multi-method fusion IQA metrics have no statistical significant different with some single-method IQA metrics such as FSIM and MAD. In the end, a thorough analysis is given to explain why the performance of multi-method fusion IQA framework drop significantly in cross database validation.
文摘This paper presents an investigation on the effect of JPEG compression on the similarity between the target image and the background,where the similarity is further used to determine the degree of clutter in the image.Four new clutter metrics based on image quality assessment are introduced,among which the Haar wavelet-based perceptual similarity index,known as HaarPSI,provides the best target acquisition prediction results.It is shown that the similarity between the target and the background at the boundary between visually lossless and visually lossy compression does not change significantly compared to the case when an uncompressed image is used.In future work,through subjective tests,it is necessary to check whether this presence of compression at the threshold of just noticeable differences will affect the human target acquisition performance.Similarity values are compared with the results of subjective tests of the well-known target Search_2 database,where the degree of agreement between objective and subjective scores,measured through linear correlation,reached a value of 90%.
基金partially supported by the Research Grants Council of the Hong Kong SAR, China (Project CUHK 415712)the Ministry of Education Academic Research Fund (AcRF) Tier 2 in Singapore under Grant No. T208B1218
文摘While quality assessment is essential for testing, optimizing, benchmarking, monitoring, and inspecting related systems and services, it also plays an essential role in the design of virtually all visual signal processing and communication algorithms, as well as various related decision-making processes. In this paper, we first provide an overview of recently derived quality assessment approaches for traditional visual signals (i.e., 2D images/videos), with highlights for new trends (such as machine learning approaches). On the other hand, with the ongoing development of devices and multimedia services, newly emerged visual signals (e.g., mobile/3D videos) are becoming more and more popular. This work focuses on recent progresses of quality metrics, which have been reviewed for the newly emerged forms of visual signals, which include scalable and mobile videos, High Dynamic Range (HDR) images, image segmentation results, 3D images/videos, and retargeted images.
基金supported by the Key Research and Development Program of Zhejiang Province(Grant No.2019C03138 and No.2019C01002)。
文摘The bandwidth of internet connections is still a bottleneck when transmitting large amounts of images,making the image quality assessment essential.Neurophysiological assessment of image quality has highlight advantages for it does not interfere with natural viewing behavior.However,in JPEG compression,the previous study is hard to tell the difference between the electroencephalogram(EEG)evoked by different quality images.In this paper,we propose an EEG analysis approach based on algebraic topology analysis,and the result shows that the difference between Euler characteristics of EEG evoked by different distortion images is striking both in the alpha and beta band.Moreover,we further discuss the relationship between the images and the EEG signals,and the results implied that the algebraic topological properties of images are consistent with that of brain perception,which is possible to give birth to braininspired image compression based on algebraic topological features.In general,an algebraic topologybased approach was proposed in this paper to analyze the perceptual characteristics of image quality,which will be beneficial to provide a reliable score for data compression in the network and improve the network transmission capacity.
基金supported in part by China"973"Program under Grant No.2014CB340303
文摘In this paper, we propose a novel image recompression frame- work and image quality assessment （IQA） method to efficiently recompress Internet images. With this framework image size is significantly reduced without affecting spatial resolution or perceptible quality of the image. With the help of IQA, the relationship between image quality and image evaluation scores can be quickly established, and the optimal quality factor can be obtained quickly and accurately within a pre - determined perceptual quality range. This process ensures the image＇s perceptual quality, which is applied to each input image. The test results show that, using the proposed method, the file size of images can be reduced by about 45%-60% without affecting their visual quality. Moreover, our new image -reeompression framework can be used in to many different application scenarios.
文摘本文提出了一种基于双交叉注意力融合的Swin-AK Transformer(Swin Transformer based on alterable kernel convolution)和手工特征相结合的智能手机拍摄图像质量评价方法。首先,提取了影响图像质量的手工特征,这些特征可以捕捉到图像中细微的视觉变化;其次,提出了Swin-AK Transformer,增强了模型对局部信息的提取和处理能力。此外,本文设计了双交叉注意力融合模块,结合空间注意力和通道注意力机制,融合了手工特征与深度特征,实现了更加精确的图像质量预测。实验结果表明,在SPAQ和LIVE-C数据集上,皮尔森线性相关系数分别达到0.932和0.885,斯皮尔曼等级排序相关系数分别达到0.929和0.858。上述结果证明了本文提出的方法能够有效地预测智能手机拍摄图像的质量。
基金supported by the National Natural Science Foundation of China(Nos.62031020 and 61771279)。
文摘Neural network methods have recently emerged as a hot topic in computed tomography(CT) imaging owing to their powerful fitting ability;however, their potential applications still need to be carefully studied because their results are often difficult to interpret and are ambiguous in generalizability. Thus, quality assessments of the results obtained from a neural network are necessary to evaluate the neural network. Assessing the image quality of neural networks using traditional objective measurements is not appropriate because neural networks are nonstationary and nonlinear. In contrast, subjective assessments are trustworthy, although they are time-and energy-consuming for radiologists. Model observers that mimic subjective assessment require the mean and covariance of images, which are calculated from numerous image samples;however, this has not yet been applied to the evaluation of neural networks. In this study, we propose an analytical method for noise propagation from a single projection to efficiently evaluate convolutional neural networks(CNNs) in the CT imaging field. We propagate noise through nonlinear layers in a CNN using the Taylor expansion. Nesting of the linear and nonlinear layer noise propagation constitutes the covariance estimation of the CNN. A commonly used U-net structure is adopted for validation. The results reveal that the covariance estimation obtained from the proposed analytical method agrees well with that obtained from the image samples for different phantoms, noise levels, and activation functions, demonstrating that propagating noise from only a single projection is feasible for CNN methods in CT reconstruction. In addition, we use covariance estimation to provide three measurements for the qualitative and quantitative performance evaluation of U-net. The results indicate that the network cannot be applied to projections with high noise levels and possesses limitations in terms of efficiency for processing low-noise projections. U-net is more effective in improving the image quality of smooth regions compared with that of the edge. LeakyReLU outperforms Swish in terms of noise reduction.
