The enlightenment of artificial intelligence large-scale model on the research of intelligent eye diagnosis in traditional Chinese medicine

Eye diagnosis is a method for inspecting systemic diseases and syndromes by observing the eyes.With the development of intelligent diagnosis in traditional Chinese medicine(TCM);artificial intelligence(AI)can improve the accuracy and efficiency of eye diagnosis.However;the research on intelligent eye diagnosis still faces many challenges;including the lack of standardized and precisely labeled data;multi-modal information analysis;and artificial in-telligence models for syndrome differentiation.The widespread application of AI models in medicine provides new insights and opportunities for the research of eye diagnosis intelli-gence.This study elaborates on the three key technologies of AI models in the intelligent ap-plication of TCM eye diagnosis;and explores the implications for the research of eye diagno-sis intelligence.First;a database concerning eye diagnosis was established based on self-su-pervised learning so as to solve the issues related to the lack of standardized and precisely la-beled data.Next;the cross-modal understanding and generation of deep neural network models to address the problem of lacking multi-modal information analysis.Last;the build-ing of data-driven models for eye diagnosis to tackle the issue of the absence of syndrome dif-ferentiation models.In summary;research on intelligent eye diagnosis has great potential to be applied the surge of AI model applications.

Prediction Model of Aircraft Icing Based on Deep Neural Network

Icing is an important factor threatening aircraft flight safety.According to the requirements of airworthiness regulations,aircraft icing safety assessment is needed to be carried out based on the ice shapes formed under different icing conditions.Due to the complexity of the icing process,the rapid assessment of ice shape remains an important challenge.In this paper,an efficient prediction model of aircraft icing is established based on the deep belief network(DBN)and the stacked auto-encoder(SAE),which are all deep neural networks.The detailed network structures are designed and then the networks are trained according to the samples obtained by the icing numerical computation.After that the model is applied on the ice shape evaluation of NACA0012 airfoil.The results show that the model can accurately capture the nonlinear behavior of aircraft icing and thus make an excellent ice shape prediction.The model provides an important tool for aircraft icing analysis.

An enhanced hybrid ensemble deep learning approach for forecasting daily PM_(2.5)

PM_(2.5) forecasting technology can provide a scientific and effective way to assist environmental governance and protect public health.To forecast PM_(2.5),an enhanced hybrid ensemble deep learning model is proposed in this research.The whole framework of the proposed model can be generalized as follows:the original PM_(2.5) series is decomposed into 8 sub-series with different frequency characteristics by variational mode decomposition(VMD);the long short-term memory(LSTM)network,echo state network(ESN),and temporal convolutional network(TCN)are applied for parallel forecasting for 8 different frequency PM_(2.5) sub-series;the gradient boosting decision tree(GBDT)is applied to assemble and reconstruct the forecasting results of LSTM,ESN and TCN.By comparing the forecasting data of the models over 3 PM_(2.5) series collected from Shenyang,Changsha and Shenzhen,the conclusions can be drawn that GBDT is a more effective method to integrate the forecasting result than traditional heuristic algorithms;MAE values of the proposed model on 3 PM_(2.5) series are 1.587,1.718 and 1.327μg/m3,respectively and the proposed model achieves more accurate results for all experiments than sixteen alternative forecasting models which contain three state-of-the-art models.

Seismic velocity inversion based on CNN-LSTM fusion deep neural network

Based on the CNN-LSTM fusion deep neural network,this paper proposes a seismic velocity model building method that can simultaneously estimate the root mean square(RMS)velocity and interval velocity from the common-midpoint(CMP)gather.In the proposed method,a convolutional neural network(CNN)Encoder and two long short-term memory networks(LSTMs)are used to extract spatial and temporal features from seismic signals,respectively,and a CNN Decoder is used to recover RMS velocity and interval velocity of underground media from various feature vectors.To address the problems of unstable gradients and easily fall into a local minimum in the deep neural network training process,we propose to use Kaiming normal initialization with zero negative slopes of rectifi ed units and to adjust the network learning process by optimizing the mean square error(MSE)loss function with the introduction of a freezing factor.The experiments on testing dataset show that CNN-LSTM fusion deep neural network can predict RMS velocity as well as interval velocity more accurately,and its inversion accuracy is superior to that of single neural network models.The predictions on the complex structures and Marmousi model are consistent with the true velocity variation trends,and the predictions on fi eld data can eff ectively correct the phase axis,improve the lateral continuity of phase axis and quality of stack section,indicating the eff ectiveness and decent generalization capability of the proposed method.

Robust multi-layer extreme learning machine using bias-variance tradeoff

As a new neural network model,extreme learning machine(ELM)has a good learning rate and generalization ability.However,ELM with a single hidden layer structure often fails to achieve good results when faced with large-scale multi-featured problems.To resolve this problem,we propose a multi-layer framework for the ELM learning algorithm to improve the model’s generalization ability.Moreover,noises or abnormal points often exist in practical applications,and they result in the inability to obtain clean training data.The generalization ability of the original ELM decreases under such circumstances.To address this issue,we add model bias and variance to the loss function so that the model gains the ability to minimize model bias and model variance,thus reducing the influence of noise signals.A new robust multi-layer algorithm called ML-RELM is proposed to enhance outlier robustness in complex datasets.Simulation results show that the method has high generalization ability and strong robustness to noise.

Multi-spacecraft Intelligent Orbit Phasing Control Considering Collision Avoidance

This paper proposes an intelligent low-thrust orbit phasing control method for multiple spacecraft by simultaneously considering fuel optimization and collision avoidance. Firstly,the minimum-fuel orbit phasing control database is generated by the indirect method associated with the homotopy technique. Then,a deep network representing the minimum-fuel solution is trained. To avoid collision for multiple spacecraft,an artificial potential function is introduced in the collision-avoidance controller. Finally,an intelligent orbit phasing control method by combining the minimum-fuel neural network controller and the collision-avoidance controller is proposed. Numerical results show that the proposed intelligent orbit phasing control is valid for the multi-satellite constellation initialization without collision.

External and Internal Validation of a Computer Assisted Diagnostic Model for Detecting Multi-Organ Mass Lesions in CT images

Objective We developed a universal lesion detector(ULDor)which showed good performance in in-lab experiments.The study aims to evaluate the performance and its ability to generalize in clinical setting via both external and internal validation.Methods The ULDor system consists of a convolutional neural network(CNN)trained on around 80 K lesion annotations from about 12 K CT studies in the DeepLesion dataset and 5 other public organ-specific datasets.During the validation process,the test sets include two parts:the external validation dataset which was comprised of 164 sets of non-contrasted chest and upper abdomen CT scans from a comprehensive hospital,and the internal validation dataset which was comprised of 187 sets of low-dose helical CT scans from the National Lung Screening Trial(NLST).We ran the model on the two test sets to output lesion detection.Three board-certified radiologists read the CT scans and verified the detection results of ULDor.We used positive predictive value(PPV)and sensitivity to evaluate the performance of the model in detecting space-occupying lesions at all extra-pulmonary organs visualized on CT images,including liver,kidney,pancreas,adrenal,spleen,esophagus,thyroid,lymph nodes,body wall,thoracic spine,etc.Results In the external validation,the lesion-level PPV and sensitivity of the model were 57.9%and 67.0%,respectively.On average,the model detected 2.1 findings per set,and among them,0.9 were false positives.ULDor worked well for detecting liver lesions,with a PPV of 78.9%and a sensitivity of 92.7%,followed by kidney,with a PPV of 70.0%and a sensitivity of 58.3%.In internal validation with NLST test set,ULDor obtained a PPV of 75.3%and a sensitivity of 52.0%despite the relatively high noise level of soft tissue on images.Conclusions The performance tests of ULDor with the external real-world data have shown its high effectiveness in multiple-purposed detection for lesions in certain organs.With further optimisation and iterative upgrades,ULDor may be well suited for extensive application to external data.

Graph-enhanced neural interactive collaborative filtering

To improve the training efficiency and recommendation accuracy in cold-start interactive recommendation systems,a new graph structure called item similarity graph is proposed on the basis of real data from a public dataset.The proposed graph is built from collaborative interactions and a deep reinforcement learning-based graph-enhanced neural interactive collaborative filtering(GE-ICF)model.The GE-ICF framework is developed with a deep reinforcement learning framework and comprises an embedding propagation layer designed with graph neural networks.Extensive experiments are conducted to investigate the efficiency of the proposed graph structure and the superiority of the proposed GE-ICF framework.Results show that in cold-start interactive recommendation systems,the proposed item similarity graph performs well in data relationship modeling,with the training efficiency showing significant improvement.The proposed GE-ICF framework also demonstrates superiority in decision modeling,thereby increasing the recommendation accuracy remarkably.

A Lightweight Temporal Convolutional Network for Human Motion Prediction

A lightweight multi-layer residual temporal convolutional network model(RTCN)is proposed to target the highly complex kinematics and temporal correlation of human motion.RTCN uses 1-D convolution to efficiently obtain the spatial structure information of human motion and extract the correlation in the time series of human motion.The residual structure is applied to the proposed network model to alleviate the problem of gradient disappearance in the deep network.Experiments on the Human 3.6M dataset demonstrate that the proposed method effectively reduces the errors of motion prediction compared with previous methods,especially of long-term prediction.

A convolutional neural artistic stylization algorithm for suppressing image distortion

Aiming at the problems of image semantic content distortion and blurred foreground and background boundaries during the transfer process of convolutional neural image stylization,we propose a convolutional neural artistic stylization algorithm for suppressing image distortion.Firstly,the VGG-19 network model is used to extract the feature map from the input content image and style image and to reconstruct the content and style.Then the transfer of the input content image and style image to the output image is constrained in the local affine transformation of the color space.And the Laplacian matting matrix is constructed by combining the local affine of the input image RGB channel.For each output blocks,affine transformation maps the RGB value of the input image to the corresponding output and position,which realizes the constraint of semantic content and the control of spatial layout.Finally,the synthesized image is superimposed on the white noise image and updated iteratively with the back propagation algorithm to minimize the loss function to complete the image stylization.Experimental results show that the method can generate images with obvious foreground and background edges,clear texture,restrained semantic content-distortion,realized spatial constraint and color mapping of the transfer images,and made the stylized images visually satisfactory.

Neighborhood fusion-based hierarchical parallel feature pyramid network for object detection

In order to improve the detection accuracy of small objects,a neighborhood fusion-based hierarchical parallel feature pyramid network(NFPN)is proposed.Unlike the layer-by-layer structure adopted in the feature pyramid network(FPN)and deconvolutional single shot detector(DSSD),where the bottom layer of the feature pyramid network relies on the top layer,NFPN builds the feature pyramid network with no connections between the upper and lower layers.That is,it only fuses shallow features on similar scales.NFPN is highly portable and can be embedded in many models to further boost performance.Extensive experiments on PASCAL VOC 2007,2012,and COCO datasets demonstrate that the NFPN-based SSD without intricate tricks can exceed the DSSD model in terms of detection accuracy and inference speed,especially for small objects,e.g.,4%to 5%higher mAP(mean average precision)than SSD,and 2%to 3%higher mAP than DSSD.On VOC 2007 test set,the NFPN-based SSD with 300×300 input reaches 79.4%mAP at 34.6 frame/s,and the mAP can raise to 82.9%after using the multi-scale testing strategy.

Interaction Energy Prediction of Organic Molecules using Deep Tensor Neural Network

The interaction energy of two molecules system plays a critical role in analyzing the interacting effect in molecular dynamic simulation.Since the limitation of quantum mechanics calculating resources,the interaction energy based on quantum mechanics can not be merged into molecular dynamic simulation for a long time scale.A deep learning framework,deep tensor neural network,is applied to predict the interaction energy of three organic related systems within the quantum mechanics level of accuracy.The geometric structure and atomic types of molecular conformation,as the data descriptors,are applied as the network inputs to predict the interaction energy in the system.The neural network is trained with the hierarchically generated conformations data set.The complex tensor hidden layers are simplified and trained in the optimization process.The predicted results of different molecular sys tems indica te that deep t ensor neural net work is capable to predic t the interaction energy with 1 kcal/mol of the mean absolute error in a relatively short time.The prediction highly improves the efficiency of interaction energy calculation.The whole proposed framework provides new insights to introducing deep learning technology into the interaction energy calculation.

Image Deraining for UAV Using Split Attention Based Recursive Network

Images captured in rainy days suffer from noticeable degradation of scene visibility.Unmanned aerial vehicles(UAVs),as important outdoor image acquisition systems,demand a proper rain removal algorithm to improve visual perception quality of captured images as well as the performance of many subsequent computer vision applications.To deal with rain streaks of different sizes and directions,this paper proposes to employ convolutional kernels of different sizes in a multi-path structure.Split attention is leveraged to enable communication across multiscale paths at feature level,which allows adaptive receptive field to tackle complex situations.We incorporate the multi-path convolution and the split attention operation into the basic residual block without increasing the channels of feature maps.Moreover,every block in our network is unfolded four times to compress the network volume without sacrificing the deraining performance.The performance on various benchmark datasets demonstrates that our method outperforms state-of-the-art deraining algorithms in both numerical and qualitative comparisons.

A deep dense captioning framework with joint localization and contextual reasoning

Dense captioning aims to simultaneously localize and describe regions-of-interest(RoIs)in images in natural language.Specifically,we identify three key problems:1)dense and highly overlapping RoIs,making accurate localization of each target region challenging;2)some visually ambiguous target regions which are hard to recognize each of them just by appearance;3)an extremely deep image representation which is of central importance for visual recognition.To tackle these three challenges,we propose a novel end-to-end dense captioning framework consisting of a joint localization module,a contextual reasoning module and a deep convolutional neural network(CNN).We also evaluate five deep CNN structures to explore the benefits of each.Extensive experiments on visual genome(VG)dataset demonstrate the effectiveness of our approach,which compares favorably with the state-of-the-art methods.

Tongue image segmentation and tongue color classification based on deep learning

Objective To propose two novel methods based on deep learning for computer-aided tongue diagnosis,including tongue image segmentation and tongue color classification,improving their diagnostic accuracy.Methods LabelMe was used to label the tongue mask and Snake model to optimize the labeling results.A new dataset was constructed for tongue image segmentation.Tongue color was marked to build a classified dataset for network training.In this research,the Inception+Atrous Spatial Pyramid Pooling(ASPP)+UNet(IAUNet)method was proposed for tongue image segmentation,based on the existing UNet,Inception,and atrous convolution.Moreover,the Tongue Color Classification Net(TCCNet)was constructed with reference to ResNet,Inception,and Triple-Loss.Several important measurement indexes were selected to evaluate and compare the effects of the novel and existing methods for tongue segmentation and tongue color classification.IAUNet was compared with existing mainstream methods such as UNet and DeepLabV3+for tongue segmentation.TCCNet for tongue color classification was compared with VGG16 and GoogLeNet.Results IAUNet can accurately segment the tongue from original images.The results showed that the Mean Intersection over Union(MIoU)of IAUNet reached 96.30%,and its Mean Pixel Accuracy(MPA),mean Average Precision(mAP),F1-Score,G-Score,and Area Under Curve(AUC)reached 97.86%,99.18%,96.71%,96.82%,and 99.71%,respectively,suggesting IAUNet produced better segmentation than other methods,with fewer parameters.Triplet-Loss was applied in the proposed TCCNet to separate different embedded colors.The experiment yielded ideal results,with F1-Score and mAP of the TCCNet reached 88.86% and 93.49%,respectively.Conclusion IAUNet based on deep learning for tongue segmentation is better than traditional ones.IAUNet can not only produce ideal tongue segmentation,but have better effects than those of PSPNet,SegNet,UNet,and DeepLabV3+,the traditional networks.As for tongue color classification,the proposed network,TCCNet,had better F1-Score and mAP values as compared with other neural networks such as VGG16 and GoogLeNet.

Research on red tide occurrence forecast method based on deep learning

As a marine disaster,red tides have a serious impact on marine fisheries,ecology,economy,human production and life.Red tides have been widely concerned by researchers for a long time.However,due to its complex formation mechanism,red tide forecasting is extremely challenging.Aiming at addressing problem of red tide forecasting,this paper collects the marine monitoring data before and after the occurrence of red tide in Xiamen sea area,and analyzes the correlation between multiple environmental factors and the red tide occurrence by combining the methods of Pearson correlation coefficient,Scatter matrix,and multiple correlation coefficient.The fusion method of LSTM and CNN based on deep learning are applied to mine the temporal dependence of environmental factors and find the local features of sequence data,then predict the occurrence of red tides.In the Xiamen No.1 and Xiamen No.2 datasets,the RMSE and MAE errors of this method are reaching 0.5218 and 0.5043,respectively.The forecast probability of red tide occurrence was further determined through the collaborative comparison model.The final forecast accuracy of the two datasets is 67.58%and 63.49%,respectively.This study provides exploratory experience for the analysis and forecasting of red tides,which proves the feasibility of applying deep learning methods to red tide forecasting.

Arrival Pattern Recognition and Prediction Based on Machine Learning

A data-driven method for arrival pattern recognition and prediction is proposed to provide air traffic controllers(ATCOs)with decision support. For arrival pattern recognition,a clustering-based method is proposed to cluster arrival patterns by control intentions. For arrival pattern prediction,two predictors are trained to estimate the most possible command issued by the ATCOs in a particular traffic situation. Training the arrival pattern predictor could be regarded as building an ATCOs simulator. The simulator can assign an appropriate arrival pattern for each arrival aircraft,just like real ATCOs do. Therefore,the simulator is considered to be able to provide effective advice for part of the work of ATCOs. Finally,a case study is carried out and demonstrates that the convolutional neural network(CNN)-based predictor performs better than the radom forest(RF)-based one.

Solving Schrodinger Equation with Soft Constrained Monotonic Neural Network

Artificial Neural Network(ANN)has become a powerful tool in the field of scientific research with its powerful information encapsulation ability and convenient variational optimization method.In particular,there have been many recent advances in computational physics to solve variational problems.Deep Neural Network(DNN)is used to represent the wave function to solve quantum many-body problems using variational optimization.In this work we used a new Physics-Informed Neural Network(PINN)to represent the Cumulative Distribution Function(CDF)of some classical problems in quantum mechanics and to obtain their ground state wave function and ground state energy through the CDF.By benchmarking against the exact solution,the error of the results can be controlled at a very low level.This new network architecture and optimization method can provide a new choice for solving quantum many-body problems.

Classification of hyperspectral images based on a convolutional neural network and spectral sensitivity

In recent years,deep learning methods have gradually come to be used in hyperspectral imaging domains.Because of the peculiarity of hyperspectral imaging,a mass of information is contained in the spectral dimensions of hyperspectral images.Also,different ob jects on a land surface are sensitive to different ranges of wavelength.To achieve higher accuracy in classification,we propose a structure that combines spectral sensitivity with a convolutional neural network by adding spectral weights derived from predicted outcomes before the final classification layer.First,samples are divided into visible light and infrared,with a portion of the samples fed into networks during training.Then,two key parameters,unrecognized rate(δ)and wrongly recognized rate(γ),are calculated from the predicted outcome of the whole scene.Next,the spectral weight,derived from these two parameters,is calculated.Finally,the spectral weight is added and an improved structure is constructed.The improved structure not only combines the features in spatial and spectral dimensions,but also gives spectral sensitivity a primary status.Compared with inputs from the whole spectrum,the improved structure attains a nearly 2%higher prediction accuracy.When applied to public data sets,compared with the whole spectrum,on the average we achieve approximately 1%higher accuracy.