Temporal Layer Research Articles

Dynamic Hand Gesture Recognition Using Improved Spatio-Temporal Graph Convolutional Network

Hand gesture recognition is essential to human-computer interaction as the most natural way of communicating. Furthermore, with the development of 3D hand pose estimation technology and the performance improvement of low-cost depth cameras, skeleton-based dynamic hand gesture recognition has received much attention. This paper proposes a novel multi-stream improved spatio-temporal graph convolutional network (MS-ISTGCN) for skeleton-based dynamic hand gesture recognition. We adopt an adaptive spatial graph convolution that can learn the relationship between distant hand joints and propose an extended temporal graph convolution with multiple dilation rates that can extract informative temporal features from short to long periods. Furthermore, we add a new attention layer consisting of effective spatio-temporal attention and channel attention between the spatial and temporal graph convolution layers to find and focus on key features. Finally, we propose a multi-stream structure that feeds multiple data modalities (i.e., joints, bones, and motions) as inputs to improve performance using the ensemble technique. Each of the three-stream networks is independently trained and fused to predict the final hand gesture. The performance of the proposed method is verified through extensive experiments with two widely used public dynamic hand gesture datasets: SHREC’17 Track and DHG-14/28. Our proposed method achieves the highest recognition accuracy in various gesture categories for both datasets compared with state-of-the-art methods.

Dynamic Spatial Transformer WaveNet Network for Traffic Forecasting

Traffic forecasting has emerged as an important task for developing intelligent transportation systems. Recent works focus on representing traffic as graph operation and using graph neural networks for spatial–temporal prediction. Most of the approaches assume a predefined graph structure based on node distances. However, spatial dependencies change over time in many scenarios of traffic flow. In this regard, this study takes an investigation capturing the spatial and temporal dependencies with no prior knowledge structure of traffic road networks. Specifically, we propose a multi-step prediction model named Dynamic Spatial Transformer WaveNet Network (DSTWN) to capture the dynamic conditions and directions of traffic flow in which a temporal convolution layer is adopted for the long time sequence and a spatial transformer layer is proposed to capture the dynamic spatial dependencies. Furthermore, we introduce a new traffic dataset, which is collected from the vehicle detection system in an urban area (UVDS). In particular, compared with existing benchmark traffic data, UVDS contains more complicated spatial information, which is similar to many real-world scenarios of traffic flow. Experiments on both benchmark traffic datasets indicate the promising results of DSTWN compared with state-of-the-art models in this research field.

Glaucoma Diagnosis Through the Integration of Optical Coherence Tomography/Angiography and Machine Learning Diagnostic Models.

PurposeTo establish optical coherence tomography (OCT)/angiography (OCTA) parameter ranges for healthy eyes (HE) and glaucomatous eyes (GE) for a North Texas based population; to develop a machine learning (ML) tool and to identify the most accurate diagnostic parameters for clinical glaucoma diagnosis.Patients and MethodsIn this retrospective cross-sectional study, we included 1371 eligible eyes, 462 HE and 909 GE (377 ocular hypertension, 160 mild, 156 moderate, 216 severe), from 735 subjects. Demographic data and full OCTA parameters were collected. A Kruskal–Wallis test was used to produce the normative database. Models were trained to solve a two-class problem (HE vs GE) and four-class problem (HE vs mild vs moderate vs severe GE). A rigorous nested, stratified, group, 5×10 fold cross-validation strategy was applied to partition the data. Six ML algorithms were compared using classical and deep learning approaches. Over 2500 ML models were optimized using random search, with performance compared using mean validation accuracy. Final performance was reported on held-out test data using accuracy and F1 score. Decision trees and feature importance were produced for the final model.ResultsWe found differences across glaucoma severities for age, gender, hypertension, Black and Asian race, and all OCTA parameters, except foveal avascular zone area and perimeter (p<0.05). The XGBoost algorithm achieved the highest test performance for both the two-class (F1 score 83.8%; accuracy 83.9%; standard deviation 0.03%) and four-class (F1 score 62.4%; accuracy 71.3%; standard deviation 0.013%) problem. A set of interpretable decision trees provided the most important predictors of the final model; inferior temporal and inferior hemisphere vessel density and peripapillary retinal nerve fiber layer thickness were identified as key diagnostic parameters.ConclusionThis study established a normative database for our North Texas based population and created ML tools utilizing OCT/A that may aid clinicians in glaucoma management.

Path Signature Neural Network of Cortical Features for Prediction of Infant Cognitive Scores.

Studies have shown that there is a tight connection between cognition skills and brain morphology during infancy. Nonetheless, it is still a great challenge to predict individual cognitive scores using their brain morphological features, considering issues like the excessive feature dimension, small sample size and missing data. Due to the limited data, a compact but expressive feature set is desirable as it can reduce the dimension and avoid the potential overfitting issue. Therefore, we pioneer the path signature method to further explore the essential hidden dynamic patterns of longitudinal cortical features. To form a hierarchical and more informative temporal representation, in this work, a novel cortical feature based path signature neural network (CF-PSNet) is proposed with stacked differentiable temporal path signature layers for prediction of individual cognitive scores. By introducing the existence embedding in path generation, we can improve the robustness against the missing data. Benefiting from the global temporal receptive field of CF-PSNet, characteristics consisted in the existing data can be fully leveraged. Further, as there is no need for the whole brain to work for a certain cognitive ability, a top K selection module is used to select the most influential brain regions, decreasing the model size and the risk of overfitting. Extensive experiments are conducted on an in-house longitudinal infant dataset within 9 time points. By comparing with several recent algorithms, we illustrate the state-of-the-art performance of our CF-PSNet (i.e., root mean square error of 0.027 with the time latency of 518 milliseconds for each sample).

Scene-Graph Augmented Data-Driven Risk Assessment of Autonomous Vehicle Decisions

There is considerable evidence that evaluating the subjective risk level of driving decisions can improve the safety of Autonomous Driving Systems (ADS) in both typical and complex driving scenarios. In this paper, we propose a novel data-driven approach that uses scene-graphs as intermediate representations for modeling the subjective risk of driving maneuvers. Our approach includes a Multi-Relation Graph Convolution Network, a Long-Short Term Memory Network, and attention layers. To train our model, we formulate subjective risk assessment as a supervised scene classification problem. We evaluate our model on both synthetic lane-changing datasets and real-driving datasets with various driving maneuvers. We show that our approach achieves a higher classification accuracy than the state-of-the-art approach on both large (96.4% vs. 91.2%) and small (91.8% vs. 71.2%) lane-changing synthesized datasets, illustrating that our approach can learn effectively even from small datasets. We also show that our model trained on a lane-changing synthesized dataset achieves an average accuracy of 87.8% when tested on a real-driving lane-changing dataset. In comparison, the state-of-the-art model trained on the same synthesized dataset only achieved 70.3% accuracy when tested on the real-driving dataset, showing that our approach can transfer knowledge more effectively. Moreover, we demonstrate that the addition of spatial and temporal attention layers improves our model’s performance and explainability. Finally, our results illustrate that our model can assess the risk of various driving maneuvers more accurately than the state-of-the-art model (86.5% vs. 58.4%, respectively).

A Deep Learning-Based Parameter Prediction Method for Coal Slime Blending Circulating Fluidized Bed Units

Coal slime blending can effectively improve the utilization rate of fossil fuels and reduce environmental pollution. However, the combustion in the furnace is unstable due to the empty pump phenomenon during the coal slurry transport. The combustion instability affects the material distribution in the furnace and harms the unit operation. The bed pressure in the circulating fluidized bed unit reflects the amount of material in the furnace. An accurate bed pressure prediction model can reflect the future material quantity in the furnace, which helps adjust the operation of the unit in a timely fashion. Thus, a deep learning-based prediction method for bed pressure is proposed in this paper. The Pearson correlation coefficient with time correction was used to screen the input variables. The Gaussian convolution kernels were used to implement the extraction of inertial delay characteristics of the data. Based on the computational theory of the temporal attention layer, the model was trained using the segmented approach. Ablation experiments verified the innovations of the proposed method. Compared with other models, the mean absolute error of the proposed model reached 0.0443 kPa, 0.0931 kPa, and 0.0345 kPa for the three data sets, respectively, which are better than those of the other models.

SiamTrans: Zero-Shot Multi-Frame Image Restoration with Pre-trained Siamese Transformers

We propose a novel zero-shot multi-frame image restoration method for removing unwanted obstruction elements (such as rains, snow, and moire patterns) that vary in successive frames. It has three stages: transformer pre-training, zero-shot restoration, and hard patch refinement. Using the pre-trained transformers, our model is able to tell the motion difference between the true image information and the obstructing elements. For zero-shot image restoration, we design a novel model, termed SiamTrans, which is constructed by Siamese transformers, encoders, and decoders. Each transformer has a temporal attention layer and several self-attention layers, to capture both temporal and spatial information of multiple frames. Only self-supervisedly pre-trained on the denoising task, SiamTrans is tested on three different low-level vision tasks (deraining, demoireing, and desnowing). Compared with related methods, SiamTrans achieves the best performances, even outperforming those with supervised learning.

Integrating nocturnal and diurnal interactions in a Neotropical pollination network

Abstract Plants establish pollination interactions with different groups of animals, including nocturnal ones that establish interactions with economically valuable and culturally important crops, as well as wild plants of conservation concern. Despite the considerable number of studies addressing the structure and dynamic of pollination networks, nocturnal interactions have been relatively overlooked. Using a multilayer network approach and considering diurnal and nocturnal interactions, we aimed to understand how interactions at different periods of the day are integrated and contribute to the network structural pattern. We also aimed to highlight how multilayer networks may give a more nuanced assessment of species importance across layers. We assembled a pollination network of an intensively studied Neotropical area by standardizing interaction data from 16 previous studies into a presence/absence (binary) network. Then, we used a multilayer network approach to evaluate the network modularity and plant species' roles in these different temporal layers. Plants were classified as nocturnal or diurnal according to the onset of floral opening and pollinators were classified according to their foraging period. The network consisted of 178 pollinator species and 158 plant species, with 870 links. Among plant species, 135 species have diurnal floral opening while 23 species are nocturnal. The multilayer network was significantly modular, and these modules differed in the composition of pollinator groups (e.g., hawk moths, bats, bees, hummingbirds), as well as of diurnal and nocturnal plants. We show that diurnal and nocturnal interactions are organized into interconnected modules in the multilayer network. Nocturnal plants had higher values of versatility and multidegree than diurnal plants, due to their role in connecting the two temporal layers. Synthesis. Our study highlights the importance of integrating different pollination systems to understand the importance of distinct components that structure pollination networks. We also illustrate the value of tapping into existing information, particularly species interaction data, from well‐studied biodiversity hotspot areas, to gain a better understanding of how communities are structured. Finally, despite the relative scarcity of nocturnal pollination network studies, we showed nocturnal plants, which often make complementary use of diurnal pollinators, to be important in connecting the temporal layers.

Logistics Optimization Strategy Based on Deep Neural Framework.

We propose a logistics optimization method based on improved graph convolutional networks to address the current problem of low product delivery rate and untimely product delivery during the peak period of e-commerce activities. Our method can learn excellent planning strategies from previous data and can give the best logistics strategy in time during the peak logistics period, which improves the product delivery rate and delivery time of logistics and greatly enhances the return on investment. First, we add a tensor rotation module to the graph convolution layer to better capture the global features of logistics nodes. Then we add inception structures in the temporal convolution layer to build multiscale temporal convolution filters to obtain temporal information of logistics nodes in different time-aware domains and reduce arithmetic power. Finally, we cooperate with e-commerce platforms to adopt logistics data as the experimental database. The experimental results show that our method greatly accelerates the logistics planning speed, improves the product delivery rate, ensures the timely delivery of products, and increases the return on investment.

Attention based spatiotemporal graph attention networks for traffic flow forecasting

Traffic flow forecasting is a crucial task in transportation and necessary for congestion mitigation, traffic control, and intelligent traffic management. Deep learning models can aid in high-accuracy traffic flow forecasting; however, the current research focuses only the ability of the model to capture dynamic spatiotemporal features, and studies on the effect of deeper network layers on spatiotemporal features—a critical factor affecting traffic flow forecasting accuracy—are limited. In this paper, we propose an attention-based spatiotemporal graph attention network (ASTGAT) model designed for network degradation and over-smoothing problems to investigate in-depth spatiotemporal information. Compared to other networks, ASTGAT can capture dynamic spatiotemporal correlations in data and deepen the network to improve prediction accuracy through multiple residual convolution and high-low feature concat. ASTGAT comprises three components that separately model the temporal relationships of the recent, daily, and weekly periods. Each component stacks multiple spatiotemporal blocks constructed using the attention mechanism, dilated gated convolution, and graph attention network. The graph and temporal attention layers capture spatiotemporal information dynamically, and the graph attention layer alleviates the over-smoothing phenomenon to deepen the network. The combined utilization of the attention mechanism and dilated gated convolution layer improves the medium and long temporal span prediction ability. We validated ASTGAT using two open highway datasets, and the results demonstrated that our ASTGAT model effectively extracts in-depth spatiotemporal information and the prediction results outperform those predicted by the current eight baselines. Our research is dedicated to establishing a better scientific basis for intelligent traffic management that can assist in decision making.

Structural retinal changes in cerebral small vessel disease

Cerebral small vessel disease (CSVD) is an important contributor to cognitive impairment and stroke. Previous research has suggested associations with alterations in single retinal layers. We have assessed changes of all individual retinal layers in CSVD using high-resolution optical coherence tomography (OCT) for the first time. Subjects with recent magnetic resonance imaging (MRI) underwent macular and peripapillary retinal imaging using OCT for this case–control study. Number and volume ratio index (WMRI) of white matter lesions (WML) were determined on MRI. Data were analyzed using multiple linear regression models. 27 CSVD patients and 9 control participants were included. Ganglion cell layer (GCL) volume was significantly reduced in patients with CSVD compared to age-matched controls (p = 0.008). In patients with CSVD, larger foveal outer plexiform layer (OPL) volume and decreased temporal peripapillary retinal nerve fiber layer (RNFL) thickness were significantly associated with a higher WMRI in linear regression when controlling for age (p ≤ 0.033). Decreased foveal GCL volume and temporal-inferior RNFL thickness at Bruch’s membrane opening (MRW), and increased temporal MRW were associated with a higher WML burden (p ≤ 0.037). Thus, we identified alterations in several OCT layers in individuals with CSVD (GCL, OPL, MRW and RNFL). Their potential diagnostic value merits further study.

A watershed water quality prediction model based on attention mechanism and Bi-LSTM.

Accurate prediction of water quality contributes to the intelligent management and control of watershed ecology. Water Quality data has time series characteristics, but the existing models only focus on the forward time series when LSTM is introduced and do not consider the effect of the reverse time series on the model. Also did not take into account the different contributions of water quality sequences to the model at different moments. In order to solve this problem, this paper proposes a watershed water quality prediction model called AT-BILSTM. The model mainly contains a Bi-LSTM layer and a temporal attention layer and introduces an attention mechanism after bidirectional feature extraction of water quality time series data to highlight the data series that have a critical impact on the prediction results. The effectiveness of the method was verified with actual datasets from four monitoring stations in Lanzhou section of the Yellow River basin in China. After comparing with the reference model, the results show that the proposed model combines the bidirectional nonlinear mapping capability of Bi-LSTM and the feature weighting feature of the attention mechanism. Taking Fuhe Bridge as an example, compared with the original LSTM model, the RMSE and MAE of the model are reduced to 0.101 and 0.059, respectively, and the R2 is improved to 0.970, which has the best prediction performance among the four cross-sections and can provide a decision basis for the comprehensive water quality management and pollutant control in the basin.

Phase Preservation Neural Network for Electroencephalography Classification in Rapid Serial Visual Presentation Task.

Neuroscience studies have demonstrated the phase-locked characteristics of some early event-related potential (ERP) components evoked by stimuli. In this study, we propose a phase preservation neural network (PPNN) to learn phase information to improve the Electroencephalography (EEG) classification in a rapid serial visual presentation (RSVP) task. The PPNN consists of three major modules that can produce spatial and temporal representations with the high discriminative ability of the EEG features for classification. We first adopt a stack of dilated temporal convolution layers to extract temporal dynamics while avoiding the loss of phase information. Considering the intrinsic channel dependence of the EEG data, a spatial convolution layer is then applied to obtain the spatial-temporal representation of the input EEG signal. Finally, a fully connected layer is adopted to extract higher-level features for the final classification. The experiments are conducted on two public and one collected EEG datasets from the RSVP task, in which we evaluated the performance and explored the capability of phase preservation of our PPNN model and visualized the extracted features. The experimental results indicate the superiority of the proposed PPNN when compared with previous methods, suggesting the PPNN is a robust model for EEG classification in RSVP task.

Improved GCN Framework for Human Motion Recognition

Human recognition models based on spatial-temporal graph convolutional neural networks have been gradually developed, and we present an improved spatial-temporal graph convolutional neural network to solve the problems of the high number of parameters and low accuracy of this type of model. The method mainly draws on the inception structure. First, the tensor rotation is added to the graph convolution layer to realize the conversion between graph node dimension and channel dimension and enhance the model’s ability to capture global information for small-scale tasks. Then the inception temporal convolution layer is added to build a multiscale temporal convolution filter to perceive temporal information under different time domains hierarchically from 4-time dimensions. It overcomes the shortcomings of temporal graph convolutional networks in the field of joint relevance of hidden layers and compensates for the information omission of small-scale graph tasks. It also limits the volume of parameters, decreases the arithmetic power, and speeds up the computation. In our experiments, we verify our model on the public dataset NTU RGB + D. Our method reduces the number of the model parameters by 50% and achieves an accuracy of 90% in the CS evaluation system and 94% in the CV evaluation system. The results show that our method not only has high recognition accuracy and good robustness in human behavior recognition applications but also has a small number of model parameters, which can effectively reduce the computational cost.

Thickened Retinal Nerve Fiber Layers Associated With High-Altitude Headache.

Purpose: This study aimed to quantify the different quadrants of the optic nerve head (ONH) and macular parameters and their changes during exposure to high altitude, and to assess their correlation with high-altitude headache (HAH). Methods: Spectral-domain optical coherence tomography (OCT) was used to quantify changes in the retinal structure in 109 healthy subjects during acute exposure to high altitude (3,700 m). Self-reported symptoms of HAH and acute mountain sickness AMS were assessed using Lake Louise Score (LLS), alongside measurements of physiological parameters (oxygen saturation [SpO2], heart rate [HR], hemoglobin level [Hb], and red blood cell [RBC] count). Measurements were taken before and after exposure to the high-altitude environment. The correlations of these parameters and changes at ONH were examined. Results: With the exposure to high altitude, the incidence of AMS was 44.0% and the frequency of HAH was 67.0% (54.1% mild, 12.9% moderate-severe). As for systemic parameters measured at high altitude, the participants exhibited significantly lower SpO2, higher resting HR, higher Hb, and a higher RBC (all p < 0.05). Key stereometric parameters used to describe ONH [superior, inferior, nasal, temporal, and mean retinal nerve fiber layer (RNFL) thickness] and macula (macular thickness) increased at high altitude compared with baseline. Most parameters of ONH changed, especially superior, inferior, and mean RNFL thickness (p < 0.05). There was a significant correlation between the ratios of RNFL at ONH and HAH [mean thickness (r = 0.246, p = 0.01); inferior (r = 0.216, p = 0.02); nasal (r = 0.193, p = 0.04)]. No associations between parameters of ONH and AMS or LLS were observed. Conclusion: The high-altitude environment can increase RNFL thickness at ONH. Furthermore, we found that the ratios of mean thickness, inferior area, and nasal area correlated positively with HAH, which provides new insights for understanding of the underlying pathological mechanisms of high-altitude retinopathy (HAR).

Spectral-domain OCT measurements in obesity: A systematic review and meta-analysis.

BackgroundPrevious studies proposed possible applications of spectral-domain optical coherence tomography (SD-OCT) measurements in prognosticating pathologies observed in overweight/obesity, including ocular, vascular, and neurologic consequences. Therefore, we conducted a systematic review and meta-analysis to investigate the changes in the in SD-OCT measurements of the patients with higher body mass index (BMI) compared to normal weight individuals.Materials and methodsWe conducted a systematic search on PubMed, Scopus, and Embase. The search results underwent two-phase title/abstract and full-text screenings. We then analyzed SD-OCT measurements differences in patients with high BMI and controls, and performed meta-regression, sub-group analysis, quality assessment, and publication bias assessment. The measurements included macular thickness, cup to disc ratio, ganglion cell-inner plexiform layer (GC-IPL) and its sub-sectors, RNFL and peripapillary RNFL (pRNFL) and their sub-layers, and choroidal thickness and its sub-sectors.Results19 studies were included in this meta-analysis accounting for 1813 individuals, 989 cases and 824 controls. There was an overall trend towards decreased thickness in high BMI patients, but only two measurements reached statistical significance: temporal retinal nerve fiber layer (RNFL) (Standardized mean difference (SMD): -0.33, 95% confidence interval (CI): -0.53 to -0.14, p<0.01) and the choroidal region 1.0 mm nasal to fovea (SMD: -0.38, 95% CI: -0.60 to -0.16, p<0.01).ConclusionSome ocular layers are thinner in patients with higher BMI than the controls. These SD-OCT measurements might correlate with adverse events related to increased body weight and have prognostic abilities. As SD-OCT is a robust, rapid and non-invasive tool, future guidelines and studies are needed to evaluate the possibility of their integration into care of the patients with obesity.

Compressible Velocity Transformations for Various Noncanonical Wall-Bounded Turbulent Flows

Although several velocity transformations for compressible zero-pressure-gradient boundary layers have been proposed in the past decades, their performance for noncanonical compressible wall-bounded turbulent flows has not been systematically investigated. This work assesses several popular transformations for the velocity profile through their application to several types of noncanonical compressible wall-bounded turbulent flows. Specifically, this work explores direct numerical simulation databases of high-enthalpy boundary layers with dissociation and vibrational excitation, supercritical channel and boundary-layer flows, and adiabatic boundary layers with pressure gradients. The transformations considered include the van Driest (“Turbulent Boundary Layer in Compressible Fluids,” Journal of the Aeronautical Sciences, Vol. 18, No. 3, 1951, pp. 145–216), Zhang et al. (“Mach-Number-Invariant Mean-Velocity Profile of Compressible Turbulent Boundary Layers,” Physical Review Letters, Vol. 109, No. 5, 2012, Paper 054502), Trettel and Larsson (“Mean Velocity Scaling for Compressible Wall Turbulence with Heat Transfer,” Physics of Fluids, Vol. 28, No. 2, 2016, Paper 026102), data-driven (Volpiani et al., “Data-Driven Compressibility Transformation for Turbulent Wall Layers,” Physical Review Fluids, Vol. 5, No. 5, 2020, Paper 052602), and total-stress-based (Griffin et al., “Velocity Transformation for Compressible Wall-Bounded Turbulent Flows with and Without Heat Transfer,” Proceedings of the National Academy of Sciences of the United States of America, Vol. 118, No. 34, 2021, Paper e2111144118) transformations. The Trettel–Larsson transformation collapses velocity profiles of high-enthalpy temporal boundary layers but not the spatial boundary layers considered. For supercritical channel flows, the Trettel–Larsson transformation also performs well over the entire inner layer. None of the aforementioned transformations work for supercritical boundary layers. For all the considered methods, the transformed velocity profiles of boundary layers with weak pressure gradients coincide well with the universal incompressible law of the wall. In summary, all these popular methods fail to deliver uniform performance for noncanonical compressible wall-bounded flows in the logarithmic region, and a more sophisticated version, which accounts for these different physics, is needed. The data-driven and total-stress-based transformations perform well in the viscous sublayer for all the considered flows. Nevertheless, the present assessment provides a useful guideline on the deployment of these transformations on various noncanonical flows.

Spatiotemporal attention mechanism-based multistep traffic volume prediction model for highway toll stations

As the fundamental part of other Intelligent Transportation Systems (ITS) applications, short-term traffic volume prediction plays an important role in various intelligent transportation tasks, such as traffic management, traffic signal control and route planning. Although Neural-network-based traffic prediction methods can produce good results, most of the models can’t be explained in an intuitive way. In this paper, we not only proposed a model that increase the short-term prediction accuracy of the traffic volume, but also improved the interpretability of the model by analyzing the internal attention score learnt by the model. we propose a spatiotemporal attention mechanism-based multistep traffic volume prediction model (SAMM). Inside the model, an LSTM-based Encoder-Decoder network with a hybrid attention mechanism is introduced, which consists of spatial attention and temporal attention. In the first level, the local and global spatial attention mechanisms considering the micro traffic evolution and macro pattern similarity, respectively, are applied to capture and amplify the features from the highly correlated entrance stations. In the second level, a temporal attention mechanism is employed to amplify the features from the time steps captured as contributing more to the future exit volume. Considering the time-dependent characteristics and the continuity of the recent evolutionary traffic volume trend, the timestamp features and historical exit volume series of target stations are included as the external inputs. An experiment is conducted using data from the highway toll collection system of Guangdong Province, China. By extracting and analyzing the weights of the spatial and temporal attention layers, the contributions of the intermediate parameters are revealed and explained with knowledge acquired by historical statistics. The results show that the proposed model outperforms the state-of-the-art model by 29.51% in terms of MSE, 13.93% in terms of MAE, and 5.69% in terms of MAPE. The effectiveness of the Encoder-Decoder framework and the attention mechanism are also verified.

Skeleton-Based Spatio-Temporal U-Network for 3D Human Pose Estimation in Video.

Despite the great progress in 3D pose estimation from videos, there is still a lack of effective means to extract spatio-temporal features of different granularity from complex dynamic skeleton sequences. To tackle this problem, we propose a novel, skeleton-based spatio-temporal U-Net(STUNet) scheme to deal with spatio-temporal features in multiple scales for 3D human pose estimation in video. The proposed STUNet architecture consists of a cascade structure of semantic graph convolution layers and structural temporal dilated convolution layers, progressively extracting and fusing the spatio-temporal semantic features from fine-grained to coarse-grained. This U-shaped network achieves scale compression and feature squeezing by downscaling and upscaling, while abstracting multi-resolution spatio-temporal dependencies through skip connections. Experiments demonstrate that our model effectively captures comprehensive spatio-temporal features in multiple scales and achieves substantial improvements over mainstream methods on real-world datasets.

Hydroxychloroquine treatment alters retinal layers and choroid without apparent toxicity in optical coherence tomography

Background To investigate whether single retinal layer, choroidal and peripapillary retinal fiber layer thicknesses were affected in patients under hydroxychloroquine treatment without retinal toxicityMethods Forty-nine patients under hydroxychloroquine treatment (HCQ group) and 39 healthy volunteers, a total of 88 individuals, were included in the study. Patients with HCQ treatment were further divided into two groups as less than 5 years of HCQ use (group 1, 29 patients) and 5 years or more of HCQ use (group 2, 20 patients). All participants underwent full ophthalmologic examination and optical coherence tomography (OCT) imaging as OCT is a recommended tool for screening toxicity by the American Association of Ophthalmology. The automatic segmentation of retinal layers included retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), retina pigment epithelium (RPE), inner retinal layers (IRL), outer retinal layers (ORL) and total retina thickness (TRT). Choroidal thickness (ChT), choroidal vascularity index and peripapillary RNFL were also measured.Results Both inferior and nasal parafoveal RNFL, temporal parafoveal GCL, and IPL layer thicknesses were decreased, while temporal parafoveal RPE layer thicknesses were increased in hydroxychloroquine users. Subfoveal ChTs were significantly reduced in both groups 1 and 2 compared to the control group(p = 0.042 and p = 0.009, respectively) and temporal ChT thicknesses were statistically decreased in group 2 compared to the control group(p = 0.018), but not for nasal ChT.Conclusions Hydroxychloroquine use may cause parafoveal retinal layer thickness alterations and also subfoveal and parafoveal ChT alterations even without evident retinal toxicity in OCT.