Automated Video Generation of Moving Digits from Text Using Deep Deconvolutional Generative Adversarial Network

Generating realistic and synthetic video from text is a highly challenging task due to the multitude of issues involved,including digit deformation,noise interference between frames,blurred output,and the need for temporal coherence across frames.In this paper,we propose a novel approach for generating coherent videos of moving digits from textual input using a Deep Deconvolutional Generative Adversarial Network(DD-GAN).The DDGAN comprises a Deep Deconvolutional Neural Network(DDNN)as a Generator(G)and a modified Deep Convolutional Neural Network(DCNN)as a Discriminator(D)to ensure temporal coherence between adjacent frames.The proposed research involves several steps.First,the input text is fed into a Long Short Term Memory(LSTM)based text encoder and then smoothed using Conditioning Augmentation(CA)techniques to enhance the effectiveness of the Generator(G).Next,using a DDNN to generate video frames by incorporating enhanced text and random noise and modifying a DCNN to act as a Discriminator(D),effectively distinguishing between generated and real videos.This research evaluates the quality of the generated videos using standard metrics like Inception Score(IS),Fréchet Inception Distance(FID),Fréchet Inception Distance for video(FID2vid),and Generative Adversarial Metric(GAM),along with a human study based on realism,coherence,and relevance.By conducting experiments on Single-Digit Bouncing MNIST GIFs(SBMG),Two-Digit Bouncing MNIST GIFs(TBMG),and a custom dataset of essential mathematics videos with related text,this research demonstrates significant improvements in both metrics and human study results,confirming the effectiveness of DD-GAN.This research also took the exciting challenge of generating preschool math videos from text,handling complex structures,digits,and symbols,and achieving successful results.The proposed research demonstrates promising results for generating coherent videos from textual input.

基于聚焦变换解卷积的宽带稳健波束形成

In the field of array signal processing,uniform linear arrays(ULAs)are widely used to detect/separate a weak target and estimate its direction of arrival from interference and noise.Conventional beamforming(CBF)is robust but restricted by a wide mainlobe and high sidelobe level.Covariance-matrix-inversed beamforming techniques,such as the minimum variance distortionless response and multiple signal classification,are sensitive to signal mismatch and data snapshots and exhibit high-resolution performance because of the narrow mainlobe and low sidelobe level.Therefore,compared with the wideband CBF,this study proposes a robust focused-and-deconvolved conventional beamforming(RFD-CBF),utilizing the Richardson–Lucy(R-L)iterative algorithm to deconvolve the focused conventional beam power of a half-wavelength spaced ULA.Then,the focused-and-deconvolved beam power achieves a narrower mainlobe and lower sidelobe level while retaining the robustness of wideband CBF.Moreover,compared with the wideband CBF,RFD-CBF can obtain a higher output signal-to-noise ratio(SNR).Finally,the performance of RFD-CBF is evaluated through numerical simulation and verified by sea trial data processing.

Distributed Dynamic Load in Structural Dynamics by the Impulse-Based Force Estimation Algorithm

This paper proposes a novel approach for identifying distributed dynamic loads in the time domain.Using polynomial andmodal analysis,the load is transformed intomodal space for coefficient identification.This allows the distributed dynamic load with a two-dimensional form in terms of time and space to be simultaneously identified in the form of modal force,thereby achieving dimensionality reduction.The Impulse-based Force Estimation Algorithm is proposed to identify dynamic loads in the time domain.Firstly,the algorithm establishes a recursion scheme based on convolution integral,enabling it to identify loads with a long history and rapidly changing forms over time.Secondly,the algorithm introduces moving mean and polynomial fitting to detrend,enhancing its applicability in load estimation.The aforementioned methodology successfully accomplishes the reconstruction of distributed,instead of centralized,dynamic loads on the continuum in the time domain by utilizing acceleration response.To validate the effectiveness of the method,computational and experimental verification were conducted.

Cryo-EM combined with image deconvolution to determine ZIF-8 crystal structure

Metal–organic frameworks(MOFs) are crystalline porous materials with tunable properties, exhibiting great potential in gas adsorption, separation and catalysis.[1,2]It is challenging to visualize MOFs with transmission electron microscopy(TEM) due to their inherent instability under electron beam irradiation. Here, we employ cryo-electron microscopy(cryoEM) to capture images of MOF ZIF-8, revealing inverted-space structural information at a resolution of up to about 1.7A and enhancing its critical electron dose to around 20 e^(-)/A^(2). In addition, it is confirmed by electron-beam irradiation experiments that the high voltage could effectively mitigate the radiolysis, and the structure of ZIF-8 is more stable along the [100] direction under electron beam irradiation. Meanwhile, since the high-resolution electron microscope images are modulated by contrast transfer function(CTF) and it is difficult to determine the positions corresponding to the atomic columns directly from the images. We employ image deconvolution to eliminate the impact of CTF and obtain the structural images of ZIF-8. As a result, the heavy atom Zn and the organic imidazole ring within the organic framework can be distinguished from structural images.

Research on a deconvolution algorithm for laser-induced fluorescence diagnosis based on the maximum entropy principle

Laser-induced fluorescence(LIF)spectroscopy is employed for plasma diagnosis,necessitating the utilization of deconvolution algorithms to isolate the Doppler effect from the raw spectral signal.However,direct deconvolution becomes invalid in the presence of noise as it leads to infinite amplification of high-frequency noise components.To address this issue,we propose a deconvolution algorithm based on the maximum entropy principle.We validate the effectiveness of the proposed algorithm by utilizing simulated LIF spectra at various noise levels(signal-to-noise ratio,SNR=20–80 d B)and measured LIF spectra with Xe as the working fluid.In the typical measured spectrum(SNR=26.23 d B)experiment,compared with the Gaussian filter and the Richardson–Lucy(R-L)algorithm,the proposed algorithm demonstrates an increase in SNR of 1.39 d B and 4.66 d B,respectively,along with a reduction in the root-meansquare error(RMSE)of 35%and 64%,respectively.Additionally,there is a decrease in the spectral angle(SA)of 0.05 and 0.11,respectively.In the high-quality spectrum(SNR=43.96 d B)experiment,the results show that the running time of the proposed algorithm is reduced by about98%compared with the R-L iterative algorithm.Moreover,the maximum entropy algorithm avoids parameter optimization settings and is more suitable for automatic implementation.In conclusion,the proposed algorithm can accurately resolve Doppler spectrum details while effectively suppressing noise,thus highlighting its advantage in LIF spectral deconvolution applications.

Enhancing the vertical resolution of lunar penetrating radar data using predictive deconvolution

The Yutu-2 rover onboard the Chang’E-4 mission performed the first lunar penetrating radar detection on the farside of the Moon.The high-frequency channel presented us with many unprecedented details of the subsurface structures within a depth of approximately 50 m.However,it was still difficult to identify finer layers from the cluttered reflections and scattering waves.We applied deconvolution to improve the vertical resolution of the radar profile by extending the limited bandwidth associated with the emissive radar pulse.To overcome the challenges arising from the mixed-phase wavelets and the problematic amplification of noise,we performed predictive deconvolution to remove the minimum-phase components from the Chang’E-4 dataset,followed by a comprehensive phase rotation to rectify phase anomalies in the radar image.Subsequently,we implemented irreversible migration filtering to mitigate the noise and diminutive clutter echoes amplified by deconvolution.The processed data showed evident enhancement of the vertical resolution with a widened bandwidth in the frequency domain and better signal clarity in the time domain,providing us with more undisputed details of subsurface structures near the Chang’E-4 landing site.

Solar image reconstruction method under atmospheric turbulence at Fuxian Lake Solar Observatory

Strong atmospheric turbulence reduces astronomical seeing,causing speckle images acquired by ground-based solar telescopes to become blurred and distorted.Severe distortion in speckle images impedes image phase deviation in the speckle masking reconstruction method,leading to the appearance of spurious imaging artifacts.Relying only on linear image degradation principles to reconstruct solar images is insufficient.To solve this problem,we propose the multiframe blind deconvolution combined with non-rigid alignment(MFBD-CNRA)method for solar image reconstruction.We consider image distortion caused by atmospheric turbulence and use non-rigid alignment to correct pixel-level distortion,thereby achieving nonlinear constraints to complement image intensity changes.After creating the corrected speckle image,we use the linear method to solve the wavefront phase,obtaining the target image.We verify the effectiveness of our method results,compared with others,using solar observation data from the 1 m new vacuum solar telescope(NVST).This new method successfully reconstructs high-resolution images of solar observations with a Fried parameter r0 of approximately 10 cm,and enhances images at high frequency.When r0 is approximately 5 cm,the new method is even more effective.It reconstructs the edges of solar graining and sunspots,and is greatly enhanced at mid and high frequency compared with other methods.Comparisons confirm the effectiveness of this method,with respect to both nonlinear and linear constraints in solar image reconstruction.This provides a suitable solution for image reconstruction in ground-based solar observations under strong atmospheric turbulence.

Deblurring,artifact-free optical coherence tomography with deconvolution-random phase modulation

Deconvolution is a commonly employed technique for enhancing image quality in optical imaging methods.Unfortu-nately,its application in optical coherence tomography(OCT)is often hindered by sensitivity to noise,which leads to ad-ditive ringing artifacts.These artifacts considerably degrade the quality of deconvolved images,thereby limiting its effect-iveness in OCT imaging.In this study,we propose a framework that integrates numerical random phase masks into the deconvolution process,effectively eliminating these artifacts and enhancing image clarity.The optimized joint operation of an iterative Richardson-Lucy deconvolution and numerical synthesis of random phase masks(RPM),termed as De-conv-RPM,enables a 2.5-fold reduction in full width at half-maximum(FWHM).We demonstrate that the Deconv-RPM method significantly enhances image clarity,allowing for the discernment of previously unresolved cellular-level details in nonkeratinized epithelial cells ex vivo and moving blood cells in vivo.

A novel method for gamma spectrum analysis of low-level and intermediate-level radioactive waste

The uncertainty of nuclide libraries in the analysis of the gamma spectra of low-and intermediate-level radioactive waste(LILW)using existing methods produces unstable results.To address this problem,a novel spectral analysis method is proposed in this study.In this method,overlapping peaks are located using a continuous wavelet transform.An improved quadratic convolution method is proposed to calculate the widths of the peaks and establish a fourth-order filter model to estimate the Compton edge baseline with the overlapping peaks.Combined with the adaptive sensitive nonlinear iterative peak,this method can effectively subtracts the background.Finally,a function describing the peak shape as a filter is used to deconvolve the energy spectrum to achieve accurate qualitative and quantitative analyses of the nuclide without the aid of a nuclide library.Gamma spectrum acquisition experiments for standard point sources of Cs-137 and Eu-152,a segmented gamma scanning experiment for a 200 L standard drum,and a Monte Carlo simulation experiment for triple overlapping peaks using the closest energy of three typical LILW nuclides(Sb-125,Sb-124,and Cs-134)are conducted.The results of the experiments indicate that(1)the novel method and gamma vision(GV)with an accurate nuclide library have the same spectral analysis capability,and the peak area calculation error is less than 4%;(2)compared with the GV,the analysis results of the novel method are more stable;(3)the novel method can be applied to the activity measurement of LILW,and the error of the activity reconstruction at the equivalent radius is 2.4%;and(4)The proposed novel method can quantitatively analyze all nuclides in LILW without a nuclide library.This novel method can improve the accuracy and precision of LILW measurements,provide key technical support for the reasonable disposal of LILW,and ensure the safety of humans and the environment.

Source Quantitative Identification by Reference-Based Cubic Blind Deconvolution Algorithm

The semi-blind deconvolution algorithm improves the separation accuracy by introducing reference information.However,the separation performance depends largely on the construction of reference signals.To improve the robustness of the semi-blind deconvolution algorithm to the reference signals and the convergence speed,the reference-based cubic blind deconvolution algorithm is proposed in this paper.The proposed algorithm can be combined with the contribution evaluation to provide trustworthy guidance for suppressing satellite micro-vibration.The normalized reference-based cubic contrast function is proposed and the validity of the new contrast function is theoretically proved.By deriving the optimal step size of gradient iteration under the new contrast function,we propose an efficient adaptive step optimization method.Furthermore,the contribution evaluation method based on vector projection is presented to implement the source contribution evaluation.Numerical simulation analysis is carried out to validate the availability and superiority of this method.Further tests given by the simulated satellite experiment and satellite ground experiment also confirm the effectiveness.The signals of control moment gyroscope and flywheel were extracted,respectively,and the contribution evaluation of vibration sources to the sensitive load area was realized.This research proposes a more accurate and robust algorithm for the source separation and provides an effective tool for the quantitative identification of the mechanical vibration sources.

IRMIRS:Inception-ResNet-Based Network for MRI Image Super-Resolution

Medical image super-resolution is a fundamental challenge due to absorption and scattering in tissues.These challenges are increasing the interest in the quality of medical images.Recent research has proven that the rapid progress in convolutional neural networks(CNNs)has achieved superior performance in the area of medical image super-resolution.However,the traditional CNN approaches use interpolation techniques as a preprocessing stage to enlarge low-resolution magnetic resonance(MR)images,adding extra noise in the models and more memory consumption.Furthermore,conventional deep CNN approaches used layers in series-wise connection to create the deeper mode,because this later end layer cannot receive complete information and work as a dead layer.In this paper,we propose Inception-ResNet-based Network for MRI Image Super-Resolution known as IRMRIS.In our proposed approach,a bicubic interpolation is replaced with a deconvolution layer to learn the upsampling filters.Furthermore,a residual skip connection with the Inception block is used to reconstruct a high-resolution output image from a low-quality input image.Quantitative and qualitative evaluations of the proposed method are supported through extensive experiments in reconstructing sharper and clean texture details as compared to the state-of-the-art methods.

Vibration Diagnosis and Optimization of Industrial Robot Based on TPA and EMD Methods

This paper proposedmethod that combined transmission path analysis(TPA)and empirical mode decomposition(EMD)envelope analysis to solve the vibration problemof an industrial robot.Firstly,the deconvolution filter timedomain TPA method is proposed to trace the source along with the time variation.Secondly,the TPA method positioned themain source of robotic vibration under typically different working conditions.Thirdly,independent vibration testing of the Rotate Vector(RV)reducer is conducted under different loads and speeds,which are key components of an industrial robot.The method of EMD and Hilbert envelope was used to extract the fault feature of the RV reducer.Finally,the structural problems of the RV reducer were summarized.The vibration performance of industrial robots was improved through the RV reducer optimization.From the whole industrial robot to the local RV Reducer and then to the internal microstructure of the reducer,the source of defect information is traced accurately.Experimental results showed that the TPA and EMD hybrid methods were more accurate and efficient than traditional time-frequency analysis methods to solve industrial robot vibration problems.

Toward real-time digital pulse process algorithms for CsI(Tl)detector array at external target facility in HIRFL-CSR

A fully digital data acquisition system based on a field-programmable gate array(FPGA) was developed for a CsI(Tl) array at the external target facility(ETF) in the Heavy Ion Research Facility in Lanzhou(HIRFL). To process the CsI(Tl) signals generated by γ-rays and light-charged ions, a scheme for digital pulse processing algorithms is proposed. Every step in the algorithms was benchmarked using standard γ and α sources. The scheme, which included a moving average filter, baseline restoration, leading-edge discrimination, moving window deconvolution, and digital charge comparison, was subsequently implemented on the FPGA. A good energy resolution of 5.7% for 1.33-MeV γ-rays and excellent α-γ identification using the digital charge comparison method were achieved, which satisfies CsI(Tl) array performance requirements.

融合ResNet18和Deconvolution的输送带横向跑偏检测方法

为精准检测输送带横向跑偏距离和倾斜角度,提出一种融合ResNet18和Deconvolution的输送带横向跑偏检测方法。首先,设计输送带横向跑偏检测系统,给出输送带横向跑偏检测流程。接着,提出融合ResNet18与Deconvolution的输送带边缘和托辊接触特征关键点提取方法,并采用热力图突出关键点,以便于计算跑偏距离及倾斜角度。最后,搭建输送带横向跑偏检测实验平台,并对该检测方法进行相关实验。实验表明,该检测方法可减弱光照、粉尘等干扰影响,平均准确率为95.17%,优于YOLOv5和Maskrcnn LSD算法,满足带式输送机输送带横向跑偏检测精度要求。

胰腺局灶性实性病变CT灌注参数特征及不同算法的比较

目的总结胰腺局灶性实性病变CT灌注参数的特征,评估Deconvolution和Maximum slope+Patlak方法所测得的灌注参数之间的一致性及两种测量方法之间是否可相互替换。方法 2015年12月至2016年11月在北京协和医院行全胰腺CT灌注检查、经术后及穿刺病理证实为胰腺癌(PAC)患者22例和胰腺神经内分泌瘤(p NET)患者22例(共37个病灶),全胰腺CT灌注检查采用管电压80kV、管电流100m A,进行28次连续动态体积扫描,静脉注射45ml碘普罗胺,速率5ml/s,随后追加40ml盐水,速率5ml/s。由1名经验丰富的放射科医师在西门子后处理工作站上分别用Maximum slope+Patlak及Deconvolution method方法进行数据分析,测量并记录其灌注参数。结果 Wilcoxon非参配对秩和检验结果显示,PAC(BFM比BFD,Z=-3.263,P=0.001;BVD比BVP,Z=-3.978,P=0.000)和p NET(BFM比BFD,Z=-5.212,P=0.000;BVD比BVP,Z=-2.633,P=0.008)两种方法所测得的灌注参数之间差异均有统计学意义。Spearman’s相关系数分析结果显示,PAC(BFM与BFD,r=0.845,P=0.000;BVD与BVP,r=0.964,P=0.000)和p NET(BFM与BFD,r=0.759,P=0.000;BVD比BVP,r=0.683,P=0.000)两种方法所测得的灌注参数间均有显著相关性。PAC的BFM/BFD和BVD/BVP几何均数及其95%一致性界限(LOA)分别为0.77(0.61~0.99)和1.42(1.13~1.79),p NET的BFM/BFD和BVD/BVP几何均数及其95%LOA分别为0.66(0.51~0.86)和1.15(0.88~1.50)。结论无论在PAC还是p NET中由Deconvolution和Maximum slope+Patlak方法所测得的灌注参数BF、BV存在显著性差异,两种测量方法之间具有显著相关性,并且两者之间的转换范围较窄,所以两种测量方法之间可以相互替换。

气相色谱-质谱数据的后处理

气相色谱-质谱(gas chromatography/mass spectrometry)以其优异的分离定性特点,被广泛地应用于分析复杂混合物中的挥发性组分.然而,其质谱图会包含一些来自于离子源污染物、柱流失物、基质干扰物所产生的离子,导致定性准确性降低.……

基于预条件共轭梯度法的盲源反褶积方法(英文)

In seismic data processing, blind deconvolution is a key technology. Introduced in this paper is a flow of one kind of blind deconvolution. The optimal precondition conjugate gradients (PCG) in Kyrlov subspace is also used to improve the stability of the algorithm. The computation amount is greatly decreased.

BLIND ADAPTIVE XPIC BASED ON HOS

This paper presents a new blind XPIC and a new adaptive blind deconvolutional algorithm based on HOS processing, which separates and equalizes the signals in real time. The simulation results demonstrate that the performance of the proposed adaptive blind algorithm,compared with the conventional algorithms, is outstanding with the feature of feasibility, stability and fast convergence rate.

Image Denoising with GAN Based Model

Image denoising is often used as a preprocessing step in computer vision tasks,which can help improve the accuracy of image processing models.Due to the imperfection of imaging systems,transmission media and recording equipment,digital images are often contaminated with various noises during their formation,which troubles the visual effects and even hinders people’s normal recognition.The pollution of noise directly affects the processing of image edge detection,feature extraction,pattern recognition,etc.,making it difficult for people to break through the bottleneck by modifying the model.Many traditional filtering methods have shown poor performance since they do not have optimal expression and adaptation for specific images.Meanwhile,deep learning technology opens up new possibilities for image denoising.In this paper,we propose a novel neural network which is based on generative adversarial networks for image denoising.Inspired by U-net,our method employs a novel symmetrical encoder-decoder based generator network.The encoder adopts convolutional neural networks to extract features,while the decoder outputs the noise in the images by deconvolutional neural networks.Specially,shortcuts are added between designated layers,which can preserve image texture details and prevent gradient explosions.Besides,in order to improve the training stability of the model,we add Wasserstein distance in loss function as an optimization.We use the peak signal-to-noise ratio(PSNR)to evaluate our model and we can prove the effectiveness of it with experimental results.When compared to the state-of-the-art approaches,our method presents competitive performance.

Potential risks of spectrum whitening deconvolution——compared with well-driven deconvolution

Deconvolution is widely used to increase the resolution of seismic data. To compare the resolution ability of conventional spectrum whitening deconvolution to thin layers with that of welldriven deconvolution, a complex sedimentary geological model was designed, and then the simulated seismic data were processed respectively by each of the two methods. The amplitude spectrum of seismic data was almost white after spectrum whitening, but the wavelet resolution was low. The amplitude spectrum after well-driven deconvolution deviated from white spectrum, but the wavelet resolution was high. Further analysis showed that if an actual reflectivity series could not well satisfy the hypothesis of white spectrum, spectrum whitening deconvolution had a potential risk of wavelet distortion, which might lead to a pitfall in high resolution seismic data interpretation. On the other hand, the wavelet after well- driven deconvolution had higher resolution both in the time and frequency domains. It is favorable for high resolution seismic interpretation and reservoir prediction.