Image super‐resolution via dynamic network

Convolutional neural networks depend on deep network architectures to extract accurate information for image super‐resolution.However,obtained information of these con-volutional neural networks cannot completely express predicted high‐quality images for complex scenes.A dynamic network for image super‐resolution(DSRNet)is presented,which contains a residual enhancement block,wide enhancement block,feature refine-ment block and construction block.The residual enhancement block is composed of a residual enhanced architecture to facilitate hierarchical features for image super‐resolution.To enhance robustness of obtained super‐resolution model for complex scenes,a wide enhancement block achieves a dynamic architecture to learn more robust information to enhance applicability of an obtained super‐resolution model for varying scenes.To prevent interference of components in a wide enhancement block,a refine-ment block utilises a stacked architecture to accurately learn obtained features.Also,a residual learning operation is embedded in the refinement block to prevent long‐term dependency problem.Finally,a construction block is responsible for reconstructing high‐quality images.Designed heterogeneous architecture can not only facilitate richer structural information,but also be lightweight,which is suitable for mobile digital devices.Experimental results show that our method is more competitive in terms of performance,recovering time of image super‐resolution and complexity.The code of DSRNet can be obtained at https://github.com/hellloxiaotian/DSRNet.

A Unified Model Fusing Region of Interest Detection and Super Resolution for Video Compression

High-resolution video transmission requires a substantial amount of bandwidth.In this paper,we present a novel video processing methodology that innovatively integrates region of interest(ROI)identification and super-resolution enhancement.Our method commences with the accurate detection of ROIs within video sequences,followed by the application of advanced super-resolution techniques to these areas,thereby preserving visual quality while economizing on data transmission.To validate and benchmark our approach,we have curated a new gaming dataset tailored to evaluate the effectiveness of ROI-based super-resolution in practical applications.The proposed model architecture leverages the transformer network framework,guided by a carefully designed multi-task loss function,which facilitates concurrent learning and execution of both ROI identification and resolution enhancement tasks.This unified deep learning model exhibits remarkable performance in achieving super-resolution on our custom dataset.The implications of this research extend to optimizing low-bitrate video streaming scenarios.By selectively enhancing the resolution of critical regions in videos,our solution enables high-quality video delivery under constrained bandwidth conditions.Empirical results demonstrate a 15%reduction in transmission bandwidth compared to traditional super-resolution based compression methods,without any perceivable decline in visual quality.This work thus contributes to the advancement of video compression and enhancement technologies,offering an effective strategy for improving digital media delivery efficiency and user experience,especially in bandwidth-limited environments.The innovative integration of ROI identification and super-resolution presents promising avenues for future research and development in adaptive and intelligent video communication systems.

RGB-guided hyperspectral image super-resolution with deep progressive learning

Due to hardware limitations,existing hyperspectral(HS)camera often suffer from low spatial/temporal resolution.Recently,it has been prevalent to super-resolve a low reso-lution(LR)HS image into a high resolution(HR)HS image with a HR RGB(or mul-tispectral)image guidance.Previous approaches for this guided super-resolution task often model the intrinsic characteristic of the desired HR HS image using hand-crafted priors.Recently,researchers pay more attention to deep learning methods with direct supervised or unsupervised learning,which exploit deep prior only from training dataset or testing data.In this article,an efficient convolutional neural network-based method is presented to progressively super-resolve HS image with RGB image guidance.Specif-ically,a progressive HS image super-resolution network is proposed,which progressively super-resolve the LR HS image with pixel shuffled HR RGB image guidance.Then,the super-resolution network is progressively trained with supervised pre-training and un-supervised adaption,where supervised pre-training learns the general prior on training data and unsupervised adaptation generalises the general prior to specific prior for variant testing scenes.The proposed method can effectively exploit prior from training dataset and testing HS and RGB images with spectral-spatial constraint.It has a good general-isation capability,especially for blind HS image super-resolution.Comprehensive experimental results show that the proposed deep progressive learning method out-performs the existing state-of-the-art methods for HS image super-resolution in non-blind and blind cases.

Numerical Simulation of Super-Resolution Structured Illumination Microscopy (SIM) Using Heintzmann-Cremer Algorithm with Non-Continuous Spatial Frequency Support

We report a comprehensive numerical study of super resolution (SR) structured illumination microscopy (SIM) utilizing the classic Heintzmann-Cremer SIM process and algorithm. In particular, we investigated the impact of the diffraction limit of the underlying imaging system on the optimal SIM grating frequency that can be used to obtain the highest SR enhancement with non-continuous spatial frequency support. Besides confirming the previous theoretical and experimental work that SR-SIM can achieve an enhancement close to 3 times the diffraction limit with grating pattern illuminations, we also observe and report a series of more subtle effects of SR-SIM with non-continuous spatial frequency support. Our simulations show that when the SIM grating frequency exceeds twice that of the diffraction limit, the higher SIM grating frequency can help achieve a higher SR enhancement for the underlying imaging systems whose diffraction limit is low, though this enhancement is obtained at the cost of losing resolution at some lower resolution targets. Our simulations also show that, for underlying imaging systems with high diffraction limits, however, SR-SIM grating frequencies above twice the diffraction limits tend to bring no significant extra enhancement. Furthermore, we observed that there exists a limit grating frequency above which the SR enhancement effect is lost, and the reconstructed images essentially have the same resolution as the one obtained directly from the underlying imaging system without using the SIM process.

Flexible,high-density,laminated ECoG electrode array for high spatiotemporal resolution foci diagnostic localization of refractory epilepsy

High spatiotemporal resolution brain electrical signals are critical for basic neuroscience research and high-precision focus diagnostic localization,as the spatial scale of some pathologic signals is at the submillimeter or micrometer level.This entails connecting hundreds or thousands of electrode wires on a limited surface.This study reported a class of flexible,ultrathin,highdensity electrocorticogram(ECoG)electrode arrays.The challenge of a large number of wiring arrangements was overcome by a laminated structure design and processing technology improvement.The flexible,ultrathin,high-density ECoG electrode array was conformably attached to the cortex for reliable,high spatial resolution electrophysiologic recordings.The minimum spacing between electrodes was 15μm,comparable to the diameter of a single neuron.Eight hundred electrodes were prepared with an electrode density of 4444 mm^(-2).In focal epilepsy surgery,the flexible,high-density,laminated ECoG electrode array with 36 electrodes was applied to collect epileptic spike waves inrabbits,improving the positioning accuracy of epilepsy lesions from the centimeter to the submillimeter level.The flexible,high-density,laminated ECoG electrode array has potential clinical applications in intractable epilepsy and other neurologic diseases requiring high-precision electroencephalogram acquisition.

High-resolution neutronics model for ^(238)Pu production in high-flux reactors

We proposed and compared three methods(filter burnup,single energy burnup,and burnup extremum analysis)to build a high-resolution neutronics model for 238Pu production in high-flux reactors.The filter burnup and single energy burnup methods have no theoretical approximation and can achieve a spectrum resolution of up to~1 eV,thereby constructing the importance curve and yield curve of the full energy range.The burnup extreme analysis method combines the importance and yield curves to consider the influence of irradiation time on production efficiency,thereby constructing extreme curves.The three curves,which quantify the transmutation rate of the nuclei in each energy region,are of physical significance because they have similar distributions.A high-resolution neutronics model for ^(238)Pu production was established based on these three curves,and its universality and feasibility were proven.The neutronics model can guide the neutron spectrum optimization and improve the yield of ^(238)Pu by up to 18.81%.The neutronics model revealed the law of nuclei transmutation in all energy regions with high spectrum resolution,thus providing theoretical support for high-flux reactor design and irradiation production of ^(238)Pu.

Rydberg-Atom Terahertz Heterodyne Receiver with Ultrahigh Spectral Resolution

Terahertz heterodyne receivers with high sensitivity and spectral resolution are crucial for various applications.Here,we present a room-temperature atomic terahertz heterodyne receiver that achieves ultrahigh sensitivity and frequency resolution.At a signal frequency of 338.7 GHz,we obtain a sensitivity of 2.88±0.09V·cm^(−1)·Hz^(−1/2) for electric field measurements.The calibrated linear dynamical range spans approximately 89 dB,ranging from−110 dBV/cm to−21 dBV/cm.We demodulate a 400 symbol stream encoded in 4-state phase-shift keying,demonstrating excellent phase detection capability.By scanning the frequency of the local oscillator,we realize a terahertz spectrometer with Hz level frequency resolution.This resolution is more than two orders of magnitude higher than that of existing terahertz spectrometers.The demonstrated terahertz heterodyne receiver holds promising potential for working across the entire terahertz spectrum,significantly advancing its practical applications.

Virtual source approach for maximizing resolution in high-penetration gamma-ray imaging

High-energy gamma-ray radiography has exceptional penetration ability and has become an indispensable nondestructive testing(NDT)tool in various fields.For high-energy photons,point projection radiography is almost the only feasible imaging method,and its spatial resolution is primarily constrained by the size of the gamma-ray source.In conventional industrial applications,gamma-ray sources are commonly based on electron beams driven by accelerators,utilizing the process of bremsstrahlung radiation.The size of the gamma-ray source is dependent on the dimensional characteristics of the electron beam.Extensive research has been conducted on various advanced accelerator technologies that have the potential to greatly improve spatial resolution in NDT.In our investigation of laser-driven gamma-ray sources,a spatial resolution of about 90μm is achieved when the areal density of the penetrated object is 120 g/cm^(2).A virtual source approach is proposed to optimize the size of the gamma-ray source used for imaging,with the aim of maximizing spatial resolution.In this virtual source approach,the gamma ray can be considered as being emitted from a virtual source within the convertor,where the equivalent gamma-ray source size in imaging is much smaller than the actual emission area.On the basis of Monte Carlo simulations,we derive a set of evaluation formulas for virtual source scale and gamma-ray emission angle.Under optimal conditions,the virtual source size can be as small as 15μm,which can significantly improve the spatial resolution of high-penetration imaging to less than 50μm.

Advancing ophthalmic delivery of flurbiprofen via synergistic chiral resolution and ion-pairing strategies

Flurbiprofen(FB),a nonsteroidal anti-inflammatory drug,is widely employed in treating ocular inflammation owing to its remarkable anti-inflammatory effects.However,the racemic nature of its commercially available formulation(Ocufen^(R))limits the full potential of its therapeutic activity,as the(S)-enantiomer is responsible for the desired antiinflammatory effects.Additionally,the limited corneal permeability of FB significantly restricts its bioavailability.In this study,we successfully separated the chiral isomers of FB to obtain the highly active(S)-FB.Subsequently,utilizing ion-pairing technology,we coupled(S)-FB with various counter-ions,such as sodium,diethylamine,trimethamine(TMA),and l-arginine,to enhance its ocular bioavailability.A comprehensive evaluation encompassed balanced solubility,octanol-water partition coefficient,corneal permeability,ocular pharmacokinetics,tissue distribution,and in vivo ocular anti-inflammatory activity of each chiral isomer salt.Among the various formulations,S-FBTMA exhibited superior water solubility(about 1–12 mg/ml),lipid solubility(1<lgP_(ow)<3)and corneal permeability.In comparison to Ocufen^(R),S-FBTMA demonstrated significantly higher in vivo antiinflammatory activity and lower ocular irritability(such as conjunctival congestion and tingling).The findings from this research highlight the potential of chiral separation and ion-pair enhanced permeation techniques in providing pharmaceutical enterprises focused on drug development with a valuable avenue for improving therapeutic outcomes.

DTAIS:Distributed trusted active identity resolution systems for the Industrial Internet

In recent years,the Industrial Internet and Industry 4.0 came into being.With the development of modern industrial intelligent manufacturing technology,digital twins,Web3 and many other digital entity applications are also proposed.These applications apply architectures such as distributed learning,resource sharing,and arithmetic trading,which make high demands on identity authentication,asset authentication,resource addressing,and service location.Therefore,an efficient,secure,and trustworthy Industrial Internet identity resolution system is needed.However,most of the traditional identity resolution systems follow DNS architecture or tree structure,which has the risk of a single point of failure and DDoS attack.And they cannot guarantee the security and privacy of digital identity,personal assets,and device information.So we consider a decentralized approach for identity management,identity authentication,and asset verification.In this paper,we propose a distributed trusted active identity resolution system based on the inter-planetary file system(IPFS)and non-fungible token(NFT),which can provide distributed identity resolution services.And we have designed the system architecture,identity service process,load balancing strategy and smart contract service.In addition,we use Jmeter to verify the performance of the system,and the results show that the system has good high concurrent performance and robustness.

Synthesizing high-resolution satellite salinity data based on multi-fractal fusion

The spaceborne platform has unprecedently provided the global eddy-permitting(typically about 0.25°)products of sea surface salinity(SSS),however the existing SSS products can hardly resolve mesoscale motions due to the heavy noises therein and the over-smoothing in denoising processes.By means of the multi-fractal fusion(MFF),the high-resolution SSS product is synthesized with the template of sea surface temperature(SST).Two low-resolution SSS products and four SST products are considered as the source data and the templates respectively to determine the best combination.The fused products are validated by the in situ observations and intercompared via SSS maps,Singularity Exponent maps and wavenumber spectra.The results demonstrate that the MFF can perform a good work in mitigating the noises and improving the resolution.The combination of the climate change initiative SSS and the remote sensing system SST can produce the 0.1°denoised product whose global mean standard derivation of salinity against Argo is 0.21 and the feature resolution can reach 30−40 km.

Fast compressed sensing spectral measurement with adaptive gradient multiscale resolution

We propose a fast,adaptive multiscale resolution spectral measurement method based on compressed sensing.The method can apply variable measurement resolution over the entire spectral range to reduce the measurement time by over 75%compared to a global high-resolution measurement.Mimicking the characteristics of the human retina system,the resolution distribution follows the principle of gradually decreasing.The system allows the spectral peaks of interest to be captured dynamically or to be specified a priori by a user.The system was tested by measuring single and dual spectral peaks,and the results of spectral peaks are consistent with those of global high-resolution measurements.

Lightweight Multi-Resolution Network for Human Pose Estimation

Human pose estimation aims to localize the body joints from image or video data.With the development of deeplearning,pose estimation has become a hot research topic in the field of computer vision.In recent years,humanpose estimation has achieved great success in multiple fields such as animation and sports.However,to obtainaccurate positioning results,existing methods may suffer from large model sizes,a high number of parameters,and increased complexity,leading to high computing costs.In this paper,we propose a new lightweight featureencoder to construct a high-resolution network that reduces the number of parameters and lowers the computingcost.We also introduced a semantic enhancement module that improves global feature extraction and networkperformance by combining channel and spatial dimensions.Furthermore,we propose a dense connected spatialpyramid pooling module to compensate for the decrease in image resolution and information loss in the network.Finally,ourmethod effectively reduces the number of parameters and complexitywhile ensuring high performance.Extensive experiments show that our method achieves a competitive performance while dramatically reducing thenumber of parameters,and operational complexity.Specifically,our method can obtain 89.9%AP score on MPIIVAL,while the number of parameters and the complexity of operations were reduced by 41%and 36%,respectively.

Improved spatiotemporal resolution of anti-scattering super-resolution label-free microscopy via synthetic wave 3D metalens imaging

Super-resolution(SR)microscopy has dramatically enhanced our understanding of biological processes.However,scattering media in thick specimens severely limits the spatial resolution,often rendering the images unclear or indistinguishable.Additionally,live-cell imaging faces challenges in achieving high temporal resolution for fast-moving subcellular structures.Here,we present the principles of a synthetic wave microscopy(SWM)to extract three-dimensional information from thick unlabeled specimens,where photobleaching and phototoxicity are avoided.SWM exploits multiple-wave interferometry to reveal the specimen’s phase information in the area of interest,which is not affected by the scattering media in the optical path.SWM achieves~0.42λ/NA resolution at an imaging speed of up to 106 pixels/s.SWM proves better temporal resolution and sensitivity than the most conventional microscopes currently available while maintaining exceptional SR and anti-scattering capabilities.Penetrating through the scattering media is challenging for conventional imaging techniques.Remarkably,SWM retains its efficacy even in conditions of low signal-to-noise ratios.It facilitates the visualization of dynamic subcellular structures in live cells,encompassing tubular endoplasmic reticulum(ER),lipid droplets,mitochondria,and lysosomes.

Ultrahigh spatiotemporal resolution beam signal reconstruction with bunch phase compensation

Various electromagnetic signals are excited by the beam in the acceleration and beam-diagnostic elements of a particle accelerator.It is important to obtain time-domain waveforms of these signals with high temporal resolution for research,such as the study of beam–cavity interactions and bunch-by-bunch parameter measurements.Therefore,a signal reconstruction algorithm with ultrahigh spatiotemporal resolution and bunch phase compensation based on equivalent sampling is proposed in this paper.Compared with traditional equivalent sampling,the use of phase compensation and setting the bunch signal zero-crossing point as the time reference can construct a more accurate reconstructed signal.The basic principles of the method,simulation,and experimental comparison are also introduced.Based on the beam test platform of the Shanghai Synchrotron Radiation Facility(SSRF)and the method of experimental verification,the factors that affect the reconstructed signal quality are analyzed and discussed,including the depth of the sampled data,quantization noise of analog-to-digital converter,beam transverse oscillation,and longitudinal oscillation.The results of the beam experiments show that under the user operation conditions of the SSRF,a beam excitation signal with an amplitude uncertainty of 2%can be reconstructed.

Revisting High-Resolution Schemes with van Albada Slope Limiter

Slope limiters play an essential role in maintaining the non-oscillatory behavior of high-resolution methods for nonlinear conservation laws.The family of minmod limiters serves as the prototype example.Here,we revisit the question of non-oscillatory behavior of high-resolution central schemes in terms of the slope limiter proposed by van Albada et al.(Astron Astrophys 108:76–84,1982).The van Albada(vA)limiter is smoother near extrema,and consequently,in many cases,it outperforms the results obtained using the standard minmod limiter.In particular,we prove that the vA limiter ensures the one-dimensional Total-Variation Diminishing(TVD)stability and demonstrate that it yields noticeable improvement in computation of one-and two-dimensional systems.

Large-field objective lens for multi-wavelength microscopy at mesoscale and submicron resolution

Conventional microscopes designed for submicron resolution in biological research are hindered by a limited field of view,typically around 1 mm.This restriction poses a challenge when attempting to simultaneously analyze various parts of a sample,such as different brain areas.In addition,conventional objective lenses struggle to perform consistently across the required range of wavelengths for brain imaging in vivo.Here we present a novel mesoscopic objective lens with an impressive field of view of 8 mm,a numerical aperture of 0.5,and a working wavelength range from 400 to 1000 nm.We achieved a resolution of 0.74μm in fluorescent beads imaging.The versatility of this lens was further demonstrated through high-quality images of mouse brain and kidney sections in a wide-field imaging system,a confocal laser scanning system,and a two-photon imaging system.This mesoscopic objective lens holds immense promise for advancing multi-wavelength imaging of large fields of view at high resolution.

The environmental analysis and site selection of mussel and large yellow croaker aquaculture areas based on high resolution remote sensing

Mussel aquaculture and large yellow croaker aquaculture areas and their environmental characteristics in Zhoushan were analyzed using satellite data and in-situ surveys.A new two-step remote sensing method was proposed and applied to determine the basic environmental characteristics of the best mussel and large yellow croaker aquaculture areas.This methodology includes the first step of extraction of the location distribution and the second step of the extraction of internal environmental factors.The fishery ranching index(FRI1,FRI2)was established to extract the mussel and the large yellow croaker aquaculture area in Zhoushan,using Gaofen-1(GF-1)and Gaofen-6(GF-6)satellite data with a special resolution of 2 m.In the second step,the environmental factors such as sea surface temperature(SST),chlorophyll a(Chl-a)concentration,current and tide,suspended sediment concentration(SSC)in mussel aquaculture area and large yellow croaker aquaculture area were extracted and analyzed in detail.The results show the following three points.(1)For the extraction of the mussel aquaculture area,FRI1 and FRI2 are complementary,and the combination of FRI1 and FRI2 is suitable to extract the mussel aquaculture area.As for the large yellow croaker aquaculture area extraction,FRI2 is suitable.(2)Mussel aquaculture and the large yellow croaker aquaculture area in Zhoushan are mainly located on the side near the islands that are away from the eastern open waters.The water environment factor template suitable for mussel and large yellow croaker aquaculture was determined.(3)This two-step remote sensing method can be used for the preliminary screening of potential site selection for the mussels and large yellow croaker aquaculture area in the future.the fishery ranching index(FRI1,FRI2)in this paper can be applied to extract the mussel and large yellow croaker aquaculture areas in coastal waters around the world.

Meta-Learning Multi-Scale Radiology Medical Image Super-Resolution

High-resolution medical images have important medical value,but are difficult to obtain directly.Limited by hardware equipment and patient’s physical condition,the resolution of directly acquired medical images is often not high.Therefore,many researchers have thought of using super-resolution algorithms for secondary processing to obtain high-resolution medical images.However,current super-resolution algorithms only work on a single scale,and multiple networks need to be trained when super-resolution images of different scales are needed.This definitely raises the cost of acquiring high-resolution medical images.Thus,we propose a multi-scale superresolution algorithm using meta-learning.The algorithm combines a metalearning approach with an enhanced depth of residual super-resolution network to design a meta-upscale module.The meta-upscale module utilizes the weight prediction property of meta-learning and is able to perform the super-resolution task of medical images at any scale.Meanwhile,we design a non-integer mapping relation for super-resolution,which allows the network to be trained under non-integer magnification requirements.Compared to the state-of-the-art single-image super-resolution algorithm on computed tomography images of the pelvic region.The meta-learning multiscale superresolution algorithm obtained a surpassing of about 2%at a smaller model volume.Testing on different parts proves the high generalizability of our algorithm.Multi-scale super-resolution algorithms using meta-learning can compensate for hardware device defects and reduce secondary harm to patients while obtaining high-resolution medical images.It can be of great use in imaging related fields.

Deep Learned Singular Residual Network for Super Resolution Reconstruction

Single image super resolution(SISR)techniques produce images of high resolution(HR)as output from input images of low resolution(LR).Motivated by the effectiveness of deep learning methods,we provide a framework based on deep learning to achieve super resolution(SR)by utilizing deep singular-residual neural network(DSRNN)in training phase.Residuals are obtained from the difference between HR and LR images to generate LR-residual example pairs.Singular value decomposition(SVD)is applied to each LR-residual image pair to decompose into subbands of low and high frequency components.Later,DSRNN is trained on these subbands through input and output channels by optimizing the weights and biases of the network.With fewer layers in DSRNN,the influence of exploding gradients is reduced.This speeds up the learning process and also improves accuracy by using skip connections.The trained DSRNN parameters yield residuals to recover the HR subbands in the testing phase.Experimental analysis shows that the proposed method results in superior performance to existingmethods in terms of subjective quality.Extensive testing results on popular benchmark datasets such as set5,set14,and urban100 for a scaling factor of 4 show the effectiveness of the proposed method across different qualitative evaluation metrics.