Images are complex multimedia data which contain rich semantic information.Most of current image description generator algorithms only generate plain description,with the lack of distinction between primary and second...Images are complex multimedia data which contain rich semantic information.Most of current image description generator algorithms only generate plain description,with the lack of distinction between primary and secondary object,leading to insufficient high-level semantic and accuracy under public evaluation criteria.The major issue is the lack of effective network on high-level semantic sentences generation,which contains detailed description for motion and state of the principal object.To address the issue,this paper proposes the Attention-based Feedback Long Short-Term Memory Network(AFLN).Based on existing codec framework,there are two independent sub tasks in our method:attention-based feedback LSTM network during decoding and the Convolutional Block Attention Module(CBAM)in the coding phase.First,we propose an attentionbased network to feedback the features corresponding to the generated word from the previous LSTM decoding unit.We implement feedback guidance through the related field mapping algorithm,which quantifies the correlation between previous word and latter word,so that the main object can be tracked with highlighted detailed description.Second,we exploit the attention idea and apply a lightweight and general module called CBAM after the last layer of VGG 16 pretraining network,which can enhance the expression of image coding features by combining channel and spatial dimension attention maps with negligible overheads.Extensive experiments on COCO dataset validate the superiority of our network over the state-of-the-art algorithms.Both scores and actual effects are proved.The BLEU 4 score increases from 0.291 to 0.301 while the CIDEr score rising from 0.912 to 0.952.展开更多
Functional MRI was used to map the brains of subjects on-line during the process of media training for the acquisition and improvement of self-regulation mechanisms. The temporal and spatial dynamics of the new neural...Functional MRI was used to map the brains of subjects on-line during the process of media training for the acquisition and improvement of self-regulation mechanisms. The temporal and spatial dynamics of the new neural network formation were studied in real and simulated (false) biofeedback game, and their qualitative characteristics were discussed. It has been shown that immersion into a virtual competitive game, controlled by physiological responses, causes a wide involvement of the cortices, characterized by a high volume of activation in the mid-temporal, occipital and frontal areas, the cuneus and the precuneus. In both forms of media training, high values of activation volume were identified in the cerebellar structures.展开更多
In order to meet the requirements of medical research,diagnosis and treatment,a new algorithm for image fusion based on the wavelet packet transform in conjunction with both subjective and objective assessments is put...In order to meet the requirements of medical research,diagnosis and treatment,a new algorithm for image fusion based on the wavelet packet transform in conjunction with both subjective and objective assessments is put forward in the paper.Compared to the wavelet transform,the wavelet packet transform is more intricate and effective for the medical image fusion.As indicated by the experimental results,parameters of the feedback system of the new algorithm are significantly superior to those of the wavelet transform,with practicability and accuracy.展开更多
Implementation of quality improvement(QI)activities is growing in health care settings across the world.Within cardiovascular imaging,areas for improvement include consistency of result reporting,greater patient safet...Implementation of quality improvement(QI)activities is growing in health care settings across the world.Within cardiovascular imaging,areas for improvement include consistency of result reporting,greater patient safety through reductions in radiation exposure,and greater effi ciency of care delivery by elimination of wasteful practices and elimination of low-value or rarely appropriate testing.QI in health care is often driven through one of several endorsed frameworks,such as Lean,Six Sigma,and the Model for Improvement.Multiple examples of how to teach QI to medical trainees and physicians have been published.In this narrative review,we explore the growth and impact of QI in cardiovascular imaging,providing specifi c examples of successful projects,barriers to conducting QI,and rewards of persistent effort toward improving care.展开更多
This paper describes a new method for active learning in content-based image retrieval. The proposed method firstly uses support vector machine (SVM) classifiers to learn an initial query concept. Then the proposed ac...This paper describes a new method for active learning in content-based image retrieval. The proposed method firstly uses support vector machine (SVM) classifiers to learn an initial query concept. Then the proposed active learning scheme employs similarity measure to check the current version space and selects images with maximum expected information gain to solicit user's label. Finally, the learned query is refined based on the user's further feedback. With the combination of SVM classifier and similarity measure, the proposed method can alleviate model bias existing in each of them. Our experiments on several query concepts show that the proposed method can learn the user's query concept quickly and effectively only with several iterations.展开更多
Based on compressive sampling transmission model, we demonstrate here a method of quality evaluation for the reconstruction images, which is promising for the transmission of unstructured signal with reduced dimension...Based on compressive sampling transmission model, we demonstrate here a method of quality evaluation for the reconstruction images, which is promising for the transmission of unstructured signal with reduced dimension. By this method, the auxiliary information of the recovery image quality is obtained as a feedback to control number of measurements from compressive sampling video stream. Therefore, the number of measurements can be easily derived at the condition of the absence of information sparsity, and the recovery image quality is effectively improved. Theoretical and experimental results show that this algorithm can estimate the quality of images effectively and is in well consistency with the traditional objective evaluation algorithm.展开更多
High frame rate and highly sensitive imaging of refractive index changes on a surface is very promising for studying the dynamics of dissolution,mixing and biological processes without the need for labeling.Here,a hig...High frame rate and highly sensitive imaging of refractive index changes on a surface is very promising for studying the dynamics of dissolution,mixing and biological processes without the need for labeling.Here,a highly sensitive distributed feedback(DFB)dye laser sensor for high frame rate imaging refractometry without moving parts is presented.DFB dye lasers are low-cost and highly sensitive refractive index sensors.The unique multi-wavelength DFB laser structure presented here comprises several areas with different grating periods.Imagingin two dimensions of space is enabled by analyzinglaser light from all areasin parallelwith an imaging spectrometer.With this multi-resonance imaging refractometry method,the spatial position in one direction is identified from the horizontal,i.e.,spectral position of the multiple laser lines which is obtained from the spectrometer charged coupled device(CCD)array.The orthogonal spatial position is obtained from the vertical spatial position on the spectrometer CCD array as in established spatially resolved spectroscopy.Here,the imaging technique is demonstrated by monitoring the motion of small sucrose molecules upon dissolution of solid sucrose in water.The omission of moving parts improves the robustness of the imaging system and allows a very high frame rate of up to 12 Hz.展开更多
A new architecture, naked-eye ghost imaging via photoelectric feedback, is developed that avoids computer algorithm processing. Instead, the proposed scheme uses a photoelectric feedback loop to first realize the corr...A new architecture, naked-eye ghost imaging via photoelectric feedback, is developed that avoids computer algorithm processing. Instead, the proposed scheme uses a photoelectric feedback loop to first realize the correlation(multiplication) process of the traditional ghost imaging system. Then, the vision persistence effect of the naked eye is exploited to implement the integral process and to generate negative images. Two kinds of feedback circuits, the digital circuit and the analog circuit, are presented that can achieve a feedback operation. Based on this design, high-contrast real-time imaging of moving objects is obtained via a special pattern-scanning architecture on a low-speed light-modulation mask.展开更多
基金This research study is supported by the National Natural Science Foundation of China(No.61672108).
文摘Images are complex multimedia data which contain rich semantic information.Most of current image description generator algorithms only generate plain description,with the lack of distinction between primary and secondary object,leading to insufficient high-level semantic and accuracy under public evaluation criteria.The major issue is the lack of effective network on high-level semantic sentences generation,which contains detailed description for motion and state of the principal object.To address the issue,this paper proposes the Attention-based Feedback Long Short-Term Memory Network(AFLN).Based on existing codec framework,there are two independent sub tasks in our method:attention-based feedback LSTM network during decoding and the Convolutional Block Attention Module(CBAM)in the coding phase.First,we propose an attentionbased network to feedback the features corresponding to the generated word from the previous LSTM decoding unit.We implement feedback guidance through the related field mapping algorithm,which quantifies the correlation between previous word and latter word,so that the main object can be tracked with highlighted detailed description.Second,we exploit the attention idea and apply a lightweight and general module called CBAM after the last layer of VGG 16 pretraining network,which can enhance the expression of image coding features by combining channel and spatial dimension attention maps with negligible overheads.Extensive experiments on COCO dataset validate the superiority of our network over the state-of-the-art algorithms.Both scores and actual effects are proved.The BLEU 4 score increases from 0.291 to 0.301 while the CIDEr score rising from 0.912 to 0.952.
文摘Functional MRI was used to map the brains of subjects on-line during the process of media training for the acquisition and improvement of self-regulation mechanisms. The temporal and spatial dynamics of the new neural network formation were studied in real and simulated (false) biofeedback game, and their qualitative characteristics were discussed. It has been shown that immersion into a virtual competitive game, controlled by physiological responses, causes a wide involvement of the cortices, characterized by a high volume of activation in the mid-temporal, occipital and frontal areas, the cuneus and the precuneus. In both forms of media training, high values of activation volume were identified in the cerebellar structures.
文摘In order to meet the requirements of medical research,diagnosis and treatment,a new algorithm for image fusion based on the wavelet packet transform in conjunction with both subjective and objective assessments is put forward in the paper.Compared to the wavelet transform,the wavelet packet transform is more intricate and effective for the medical image fusion.As indicated by the experimental results,parameters of the feedback system of the new algorithm are significantly superior to those of the wavelet transform,with practicability and accuracy.
文摘Implementation of quality improvement(QI)activities is growing in health care settings across the world.Within cardiovascular imaging,areas for improvement include consistency of result reporting,greater patient safety through reductions in radiation exposure,and greater effi ciency of care delivery by elimination of wasteful practices and elimination of low-value or rarely appropriate testing.QI in health care is often driven through one of several endorsed frameworks,such as Lean,Six Sigma,and the Model for Improvement.Multiple examples of how to teach QI to medical trainees and physicians have been published.In this narrative review,we explore the growth and impact of QI in cardiovascular imaging,providing specifi c examples of successful projects,barriers to conducting QI,and rewards of persistent effort toward improving care.
文摘This paper describes a new method for active learning in content-based image retrieval. The proposed method firstly uses support vector machine (SVM) classifiers to learn an initial query concept. Then the proposed active learning scheme employs similarity measure to check the current version space and selects images with maximum expected information gain to solicit user's label. Finally, the learned query is refined based on the user's further feedback. With the combination of SVM classifier and similarity measure, the proposed method can alleviate model bias existing in each of them. Our experiments on several query concepts show that the proposed method can learn the user's query concept quickly and effectively only with several iterations.
基金Supported by the National Natural Science Foundation of China (No. 60972039)Jiangsu Province Natural Science Fund Project (BK2010077)Innovation Project of SCI & Tech for College Graduates of Jiangsu Province(CXLX12 _0475)
文摘Based on compressive sampling transmission model, we demonstrate here a method of quality evaluation for the reconstruction images, which is promising for the transmission of unstructured signal with reduced dimension. By this method, the auxiliary information of the recovery image quality is obtained as a feedback to control number of measurements from compressive sampling video stream. Therefore, the number of measurements can be easily derived at the condition of the absence of information sparsity, and the recovery image quality is effectively improved. Theoretical and experimental results show that this algorithm can estimate the quality of images effectively and is in well consistency with the traditional objective evaluation algorithm.
基金The authors thank JS Clausen for assistance with electron beam lithography and CLC Smith,KT Sørensen and E Højlund-Nielsen for fruitful discussionsCV acknowledges support from the Danish Research Council for Technology and Production Sciences(Grant No.12-126676).
文摘High frame rate and highly sensitive imaging of refractive index changes on a surface is very promising for studying the dynamics of dissolution,mixing and biological processes without the need for labeling.Here,a highly sensitive distributed feedback(DFB)dye laser sensor for high frame rate imaging refractometry without moving parts is presented.DFB dye lasers are low-cost and highly sensitive refractive index sensors.The unique multi-wavelength DFB laser structure presented here comprises several areas with different grating periods.Imagingin two dimensions of space is enabled by analyzinglaser light from all areasin parallelwith an imaging spectrometer.With this multi-resonance imaging refractometry method,the spatial position in one direction is identified from the horizontal,i.e.,spectral position of the multiple laser lines which is obtained from the spectrometer charged coupled device(CCD)array.The orthogonal spatial position is obtained from the vertical spatial position on the spectrometer CCD array as in established spatially resolved spectroscopy.Here,the imaging technique is demonstrated by monitoring the motion of small sucrose molecules upon dissolution of solid sucrose in water.The omission of moving parts improves the robustness of the imaging system and allows a very high frame rate of up to 12 Hz.
基金supported by the Shaanxi Key Research and Development Project (No. 2019ZDLGY09-10)the Key Scientific and Technological Innovation Team of Shaanxi Province (No. 2018TD-024)+1 种基金the National Basic Research Program of China (973 Program)(No. 2015CB654602)the 111 Project of China(No. B14040)。
文摘A new architecture, naked-eye ghost imaging via photoelectric feedback, is developed that avoids computer algorithm processing. Instead, the proposed scheme uses a photoelectric feedback loop to first realize the correlation(multiplication) process of the traditional ghost imaging system. Then, the vision persistence effect of the naked eye is exploited to implement the integral process and to generate negative images. Two kinds of feedback circuits, the digital circuit and the analog circuit, are presented that can achieve a feedback operation. Based on this design, high-contrast real-time imaging of moving objects is obtained via a special pattern-scanning architecture on a low-speed light-modulation mask.
