Spatiotemporal Relationships Research Articles

Spatio-temporal prediction of total energy consumption in multiple regions using explainable deep neural network

Energy consumption forecasting is essential for energy system integration and management. However, existing studies mainly focus on temporal features of energy consumption, which neglects the spatial correlation of variables with time information. Capturing the spatio-temporal relationships helps to improve forecasting accuracy and further promote energy dispatch. To tackle this problem, an explainable Convolutional Neural Network-Long Short Term Memory forecasting model is employed to effectively predict the total energy consumption by capturing the spatial and temporal features of multivariate time series. In the model, the autoencoder is used to achieve the nonlinear dimensionality reduction and transfer the data to a low-dimensional space. Furthermore, a Convolutional Neural Network is used to extract more effective features from the decoded data, and long short-term memory is employed to identify the temporal dependencies between extracted features and total energy consumption. Shapley additive explanation is introduced to interpret the outputs of the black-box model. The superior performance of the proposed method with high accuracy and good adaptability is verified by the comparisons with conventional forecasting models. This method provides an insight into the regional energy consumption analyzing contributions of weather variables to energy consumption, which helps administers in understanding regional energy performance for enhancing energy efficiency.

Spatiotemporal analysis of the association between residential eviction and fatal overdose in Rhode Island

ObjectivePolicy ramifications of the COVID-19 pandemic shape the concurrent housing and overdose crises in the USA. Housing insecurity is a known risk factor for overdose, yet how residential eviction may...

TransEBUS: The interpretation of endobronchial ultrasound image using hybrid transformer for differentiating malignant and benign mediastinal lesions

The purpose of this study is to establish a deep learning automatic assistance diagnosis system for benign and malignant classification of mediastinal lesions in endobronchial ultrasound (EBUS) images. EBUS images are in the form of video and contain multiple imaging modes. Different imaging modes and different frames can reflect the different characteristics of lesions. Compared with previous studies, the proposed model can efficiently extract and integrate the spatiotemporal relationships between different modes and does not require manual selection of representative frames. In recent years, Vision Transformer has received much attention in the field of computer vision. Combined with convolutional neural networks, hybrid transformers can also perform well on small datasets. This study designed a novel deep learning architecture based on hybrid transformer called TransEBUS. By adding learnable parameters in the temporal dimension, TransEBUS was able to extract spatiotemporal features from insufficient data. In addition, we designed a two-stream module to integrate information from three different imaging modes of EBUS. Furthermore, we applied contrastive learning when training TransEBUS, enabling it to learn discriminative representation of benign and malignant mediastinal lesions. The results show that TransEBUS achieved a diagnostic accuracy of 82% and an area under the curve of 0.8812 in the test dataset, outperforming other methods. It also shows that several models can improve performance by incorporating two-stream module. Our proposed system has shown its potential to help physicians distinguishing benign and malignant mediastinal lesions, thereby ensuring the accuracy of EBUS examination.

A Cold Breath of Wind: The Storm as Archipelagic Epistemology in the Horn of Africa

The Horn of Africa, located on the western edge of the Indian Ocean, is a zone of extreme weather. Often figuring within discourses of geopolitical fragmentation and environmental disaster, the Horn is also home to rich literary traditions that bear witness to its long history as a site of cultural exchange and imperial ambition. This article presents a close reading of Igabia Scego's Italian-language novel Adua (2015), arguing that a localized iteration of the Indian Ocean monsoon, the rainstorm, organizes the novel's narrative, structure, and epistemology. This “storm form” embedded in the novel is a mode of archipelagic thinking that indexes situated African environmental epistemes and the climatological and embodied effects of empire while dispelling reductive or romanticized representations. As an alternate mode of organizing spatiotemporal relations and a method of inquiry into them, the storm is a situated Horn of Africa contribution to the theorization of the Indian Ocean as monsoonal archipelago.

Spatio-Temporal Graph Representation Learning for Fraudster Group Detection.

Motivated by potential financial gain, companies may hire fraudster groups to write fake reviews to either demote competitors or promote their own businesses. Such groups are considerably more successful in misleading customers, as people are more likely to be influenced by the opinion of a large group. To detect such groups, a common model is to represent fraudster groups' static networks, consequently overlooking the longitudinal behavior of a reviewer, thus, the dynamics of coreview relations among reviewers in a group. Hence, these approaches are incapable of excluding outlier reviewers, which are fraudsters intentionally camouflaging themselves in a group and genuine reviewers happen to coreview in fraudster groups. To address this issue, we propose "FGDT," a framework for "fraudster group detection through temporal relations." FGDT first capitalizes on the effectiveness of the HIN-recurrent neural network (RNN) in both reviewers' representation learning while capturing the collaboration between reviewers. The HIN-RNN models the coreview relations of reviewers in a group in a fixed time window of 28 days. We refer to this as spatial relation learning representation to signify the generalizability of this work to other networked scenarios. Then, we use an RNN on the spatial relations to predict the spatio-temporal relations of reviewers in the group. In the third step, a graph convolution network (GCN) refines the reviewers' vector representations using these predicted relations. These refined representations are then used to remove outlier reviewers. The average of the remaining reviewers' representation is then fed to a simple fully connected layer to predict if the group is a fraudster group or not. Exhaustive experiments of FGDT showed a 5% (4%), 12% (5%), and 12% (5%) improvement over three of the most recent approaches on precision, recall, and F1-value over the Yelp (Amazon) dataset, respectively.

Aberrant cortical activity, functional connectivity, and neural assembly architecture after photothrombotic stroke in mice.

Despite substantial progress in mapping the trajectory of network plasticity resulting from focal ischemic stroke, the extent and nature of changes in neuronal excitability and activity within the peri-infarct cortex of mice remains poorly defined. Most of the available data have been acquired from anesthetized animals, acute tissue slices, or infer changes in excitability from immunoassays on extracted tissue, and thus may not reflect cortical activity dynamics in the intact cortex of an awake animal. Here, in vivo two-photon calcium imaging in awake, behaving mice was used to longitudinally track cortical activity, network functional connectivity, and neural assembly architecture for 2 months following photothrombotic stroke targeting the forelimb somatosensory cortex. Sensorimotor recovery was tracked over the weeks following stroke, allowing us to relate network changes to behavior. Our data revealed spatially restricted but long-lasting alterations in somatosensory neural network function and connectivity. Specifically, we demonstrate significant and long-lasting disruptions in neural assembly architecture concurrent with a deficit in functional connectivity between individual neurons. Reductions in neuronal spiking in peri-infarct cortex were transient but predictive of impairment in skilled locomotion measured in the tapered beam task. Notably, altered neural networks were highly localized, with assembly architecture and neural connectivity relatively unaltered a short distance from the peri-infarct cortex, even in regions within 'remapped' forelimb functional representations identified using mesoscale imaging with anaesthetized preparations 8 weeks after stroke. Thus, using longitudinal two-photon microscopy in awake animals, these data show a complex spatiotemporal relationship between peri-infarct neuronal network function and behavioral recovery. Moreover, the data highlight an apparent disconnect between dramatic functional remapping identified using strong sensory stimulation in anaesthetized mice compared to more subtle and spatially restricted changes in individual neuron and local network function in awake mice during stroke recovery.

Aberrant cortical activity, functional connectivity, and neural assembly architecture after photothrombotic stroke in mice

Despite substantial progress in mapping the trajectory of network plasticity resulting from focal ischemic stroke, the extent and nature of changes in neuronal excitability and activity within the peri-infarct cortex of mice remains poorly defined. Most of the available data have been acquired from anesthetized animals, acute tissue slices, or infer changes in excitability from immunoassays on extracted tissue, and thus may not reflect cortical activity dynamics in the intact cortex of an awake animal. Here, in vivo two-photon calcium imaging in awake, behaving mice was used to longitudinally track cortical activity, network functional connectivity, and neural assembly architecture for 2 months following photothrombotic stroke targeting the forelimb somatosensory cortex. Sensorimotor recovery was tracked over the weeks following stroke, allowing us to relate network changes to behavior. Our data revealed spatially restricted but long-lasting alterations in somatosensory neural network function and connectivity. Specifically, we demonstrate significant and long-lasting disruptions in neural assembly architecture concurrent with a deficit in functional connectivity between individual neurons. Reductions in neuronal spiking in peri-infarct cortex were transient but predictive of impairment in skilled locomotion measured in the tapered beam task. Notably, altered neural networks were highly localized, with assembly architecture and neural connectivity relatively unaltered a short distance from the peri-infarct cortex, even in regions within ‘remapped’ forelimb functional representations identified using mesoscale imaging with anaesthetized preparations 8 weeks after stroke. Thus, using longitudinal two-photon microscopy in awake animals, these data show a complex spatiotemporal relationship between peri-infarct neuronal network function and behavioral recovery. Moreover, the data highlight an apparent disconnect between dramatic functional remapping identified using strong sensory stimulation in anaesthetized mice compared to more subtle and spatially restricted changes in individual neuron and local network function in awake mice during stroke recovery.

Spatio-temporal relation of urban density and surface urban heat island index

Spatio-temporal relation of urban density and surface urban heat island index

Riverine sediment response to deforestation in the Amazon basin

Abstract. The Amazon experiences thousands of square kilometers of deforestation annually with recent rates increasing to levels unseen since the late 2000s. These increased rates of deforestation within the basin have led to changes in sediment concentration within its river systems, with potential impacts on ecological functioning, freshwater availability, and fluvial and coastal geomorphic processes. The relationship between deforestation and fluvial sediment dynamics in the Amazon has not been extensively studied using a basin-wide, comparative approach primarily due to lack of data. In this study, we utilize a novel remote-sensing-derived sediment concentration dataset to analyze the impact of deforestation from 2001 to 2020 on suspended sediment in large rivers (>50 m wide) across the Amazon River basin. These impacts are studied using a lag-based approach to quantify the spatiotemporal relationships between observed suspended sediment and changes in land cover over time. The results show that large-scale deforestation of the Amazon during the 2001–2020 period are associated with significant changes in sediment concentration in the eastern portion of the basin. In the heavily deforested eastern regions, the hydrogeomorphic response to deforestation occurs relatively rapidly (within a year), whereas the less disturbed western areas exhibit delays of 1 to 2 years before responses are observable. Moreover, we observe that deforestation must be substantial enough to overcome the collective influences of human activities and natural sediment variations to result in a discernible impact on sediment concentration in large rivers. In 69 % of Amazonian major tributary basins with an immediate response, more than 5 % of the basin was deforested during the 2001–2020 period, while in 85 % of basins with lagged responses, less than 5 % of the land was cleared. These findings suggest severe implications for future sediment dynamics across the Amazon if deforestation is to further expand into the basin.

Matching Compound Prototypes for Few-Shot Action Recognition

The task of few-shot action recognition aims to recognize novel action classes using only a small number of labeled training samples. How to better describe the action in each video and how to compare the similarity between videos are two of the most critical factors in this task. Directly describing the video globally or by its individual frames cannot well represent the spatiotemporal dependencies within an action. On the other hand, naively matching the global representations of two videos is also not optimal since action can happen at different locations in a video with different speeds. In this work, we propose a novel approach that describes each video using multiple types of prototypes and then computes the video similarity with a particular matching strategy for each type of prototypes. To better model the spatiotemporal dependency, we describe the video by generating prototypes that model the multi-level spatiotemporal relations via transformers. There are a total of three types of prototypes. The first type of prototypes are trained to describe specific aspects of the action in the video e.g., the start of the action, regardless of its timestamp. These prototypes are directly matched one-to-one between two videos to compare their similarity. The second type of prototypes are the timestamp-centered prototypes that are trained to focus on specific timestamps of the video. To deal with the temporal variation of actions in a video, we apply bipartite matching to allow the matching of prototypes of different timestamps. The third type of prototypes are generated from the timestamp-centered prototypes, which regularize their temporal consistency while serving as an auxiliary summarization of the whole video. Experiments demonstrate that our proposed method achieves state-of-the-art results on multiple benchmarks.

Comprehensive review of Transformer‐based models in neuroscience, neurology, and psychiatry

AbstractThis comprehensive review aims to clarify the growing impact of Transformer‐based models in the fields of neuroscience, neurology, and psychiatry. Originally developed as a solution for analyzing sequential data, the Transformer architecture has evolved to effectively capture complex spatiotemporal relationships and long‐range dependencies that are common in biomedical data. Its adaptability and effectiveness in deciphering intricate patterns within medical studies have established it as a key tool in advancing our understanding of neural functions and disorders, representing a significant departure from traditional computational methods. The review begins by introducing the structure and principles of Transformer architectures. It then explores their applicability, ranging from disease diagnosis and prognosis to the evaluation of cognitive processes and neural decoding. The specific design modifications tailored for these applications and their subsequent impact on performance are also discussed. We conclude by providing a comprehensive assessment of recent advancements, prevailing challenges, and future directions, highlighting the shift in neuroscientific research and clinical practice towards an artificial intelligence‐centric paradigm, particularly given the prominence of Transformer architecture in the most successful large pre‐trained models. This review serves as an informative reference for researchers, clinicians, and professionals who are interested in understanding and harnessing the transformative potential of Transformer‐based models in neuroscience, neurology, and psychiatry.

Spatio-temporal relationship between three-dimensional deformations of a collapsible tube and the downstream flowfield

The interactions between fluid flow and structural components of collapsible tubes are representative of those in several physiological systems. Although extensively studied, there exists a lack of characterization of the three-dimensionality in the structural deformations of the tube and its influence on the flow field. This experimental study investigates the spatio-temporal relationship between 3D tube geometry and the downstream flow field under conditions of fully open, closed, and slamming-type oscillating regimes. A methodology is implemented to simultaneously measure three-dimensional surface deformations in a collapsible tube and the corresponding downstream flow field. Stereophotogrammetry was used to measure tube deformations, and simultaneous flow field measurements included pressure and planar Particle Image Velocimetry (PIV) data downstream of the collapsible tube. The results indicate that the location of the largest collapse in the tube occurs close to the downstream end. In the oscillating regime, sections of the tube downstream of the largest mean collapse experience the largest oscillations in the entire tube that are completely coherent and in phase. At a certain streamwise distance upstream of the largest collapse, a switch in the direction of oscillations occurs with respect to those downstream. Physically, when the tube experiences constriction downstream of the location of the largest mean collapse, this causes the accumulation of fluid and build-up of pressure in the upstream regions and an expansion of these sections. Fluctuations in the downstream flow field are significantly influenced by tube fluctuations along the minor axes. The fluctuations in the downstream flowfield are influenced by the propagation of disturbances due to oscillations in tube geometry, through the advection of fluid through the tube. Further, the manifestation of the LU-type pressure fluctuations is found to be due to the variation in the propagation speed of the disturbances during the different stages within a period of oscillation of the tube.

Does gentrification precede and follow greening? Evidence about the green gentrification cycle in Los Angeles and Chicago

The health benefits of green space have led to calls for equitable access to parks. When new green spaces are built in low-income communities, however, gentrification often ensues. The “green gentrification” literature has paid little attention to gentrification that might occur before greening. In this paper, we explore whether and under which circumstances gentrification might precede and follow greening, a process known as the “green gentrification cycle.” Focusing on Los Angeles and Chicago, we build hedonic models focused on single-family homes sold in 2010–2021 and run post-hoc pairwise comparison tests to assess price changes over time for homes within a half-mile of new parks (treatment) and farther (control). When considering all new parks, we find that gentrification is associated with subsequent greening and, to some extent, that greening is associated with ensuing gentrification in Los Angeles. In Chicago, gentrification is associated with subsequent greening, but new parks are not associated with ensuing gentrification. Models distinguishing larger regional from smaller non-regional parks show evidence of gentrification before and after non-regional parks were built in Los Angeles, but not for regional parks. In Los Angeles, we identify stronger evidence of gentrification preceding greening farther from downtown than closer to downtown, and signals of gentrification after greening near downtown but not farther away. Models focusing on new parks built near transit stations show inconclusive results. Despite methodological limitations (e.g., examining a single measure of gentrification and different park types together), our findings shed light on the complex spatiotemporal relationships between greening and gentrification.

A reasoning method for rice fertilization strategy based on spatiotemporal knowledge graph

AbstractThe lack of multidimensional knowledge means that current reasoning methods for rice fertilization cannot make decisions accurate when faced with complex spatiotemporal conditions in general. Therefore, we propose a reasoning method for rice fertilization strategy based on spatiotemporal knowledge graph. First, we systematically organize multisource expert knowledge about rice fertilization, and construct an ontology for rice fertilization consisting of five core elements: rice variety, planting environment, nutrition diagnosis, fertilization schemes, and time and place. Spatiotemporal differences in rice fertilization knowledge are expressed by assessing spatiotemporal concepts, relations, and state instances. Second, we propose a reasoning method for rice fertilization strategy based on the constructed knowledge graph. This method leverages a certainty factor model for nutrition diagnosis and integrates case‐based and rule‐based reasoning to determine fertilization schemes for different stages. Finally, taking Pucheng County, China, as an example, knowledge from crowd‐sensing data is obtained to construct a knowledge graph using the proposed method. The results demonstrate the method can support the expression and complex reasoning of rice fertilization decisions under different spatiotemporal conditions.

Spatiotemporal Heterogeneous Responses of Ecosystem Services to Landscape Patterns in Urban–Suburban Areas

With the acceleration of urbanization, the ecosystem around cities is facing severe challenges. The drastic changes in the landscape pattern, especially in urban–suburban areas, are usually regarded as one of the main drivers. However, the spatiotemporal heterogeneous impacts of landscape patterns on the ecosystem services in this region remain unclear. To address this issue, we propose a novel framework integrating the InVEST-based ecosystem service assessment and spatiotemporal weighted regression (STWR)-based analysis of the spatiotemporal heterogeneity in urban–suburban areas, and apply it to the empirical study of Fuzhou City from 2000 to 2020. It first utilized the InVEST model to build a comprehensive ecosystem service index (CES) from five aspects (i.e., habitat quality, carbon storage, water yield, soil retention, and water purification capacity). Then, four landscape pattern indices (LPIs) (i.e., patch density (PD), area-weighted mean fractal dimension (FRAC_AM), splitting (SPLIT), and Shannon’s diversity (SHDI) index) were selected to build the STWR model. We compared and analyzed the differences in the spatial coefficient surfaces and significance tests generated by the STWR model in urban, urban–suburban, and rural areas. Results show that the following: (1) The CES in Fuzhou shows an upward trend from the urban area to the urban–suburban and rural areas, with significant gradient differences. (2) Compared with other areas, the LPIs in urban–suburban areas show more fragmentation, discreteness, and diversity, indicating more socioeconomic activities. (3) Although LPIs’ impacts on CES change over time (increasing from 2005 to 2010 and 2020 but decreasing in 2015), their effects are relatively low in urban–suburban areas, significantly lower than in urban areas. (4) Interestingly, the LPI coefficients near the urban–suburban boundary seem more significant. (5) This framework can effectively reveal the spatiotemporal heterogeneous relationships between various LPIs and CES, thus guiding concrete policies and measures that support decision-making for improving the ecosystem services surrounding cities through shaping landscape patterns.

PDED-ConvLSTM: Pyramid Dilated Deeper Encoder–Decoder Convolutional LSTM for Arctic Sea Ice Concentration Prediction

Arctic sea ice concentration plays a key role in the global ecosystem. However, accurate prediction of Arctic sea ice concentration remains a challenging task due to its inherent nonlinearity and complex spatiotemporal correlations. To address these challenges, we propose an innovative encoder–decoder pyramid dilated convolutional long short-term memory network (DED-ConvLSTM). The model is constructed based on the convolutional long short-term memory network (ConvLSTM) and, for the first time, integrates the encoder–decoder architecture of ConvLSTM (ED-ConvLSTM) with a pyramidal dilated convolution strategy. This approach aims to efficiently capture the spatiotemporal properties of the sea ice concentration and to enhance the identification of its nonlinear relationships. By applying convolutional layers with different dilation rates, the PDED-ConvLSTM model can capture spatial features at multiple scales and increase the receptive field without losing resolution. Further, the integration of the pyramid convolution module significantly enhances the model’s ability to understand complex spatiotemporal relationships, resulting in notable improvements in prediction accuracy and generalization ability. The experimental results show that the sea ice concentration distribution predicted by the PDED-ConvLSTM model is in high agreement with ground-based observations, with the residuals between the predictions and observations maintained within a range from −20% to 20%. PDED-ConvLSTM outperforms other models in terms of prediction performance, reducing the RMSE by 3.6% compared to the traditional ConvLSTM model and also performing well over a five-month prediction period. These experiments demonstrate the potential of PDED-ConvLSTM in predicting Arctic sea ice concentrations, making it a viable tool to meet the requirements for accurate prediction and provide technical support for safe and efficient operations in the Arctic region.

Enhanced stock price forecasting through a regularized ensemble framework with graph convolutional networks

Stock prediction is a prominent and challenging topic of interest for both investors and researchers. With the introduction of Graph Convolutional Networks (GCNs) to financial data analysis, many researchers have attempted to use GCNs to exploit the interrelationships among stock data. However, a notable research gap remains in the application of GCNs to establish associations among transaction data, which inherently exhibit more stable and well-defined correlations. Hence, this study proposes a regularized ensemble framework with graph convolutional networks for stock price prediction. The framework employs multiple multi-graph convolutional networks to construct an ensemble graph convolutional network, elucidating spatiotemporal relationships within transaction data across various time scales. Additionally, the Variational Modal Decomposition (VMD) method is utilized to extract features from each dimensional sequence of the transaction data, enabling the capture of features across different time scales and mitigating non-stationarity and noise in the data. Furthermore, a regularization term incorporating trend accuracy is introduced to strike a balance between numerical precision and trend accuracy, thereby enhancing the overall model performance. Finally, this study presents a specialized investment strategy designed for stock price prediction. Our approach yields substantial improvements on SSE and DJIA datasets, with ACC and R2 improving by over 15% and 10%, respectively, accompanied by reductions of more than 40% in RMSE, MAE, and MAPE compared to the baseline. These results underscore the effectiveness, robustness, and adaptability of our proposed methodology. Our codes and data can be accessed at https://github.com/WonderNothing/REGCN/.

Anomaly detection of wind turbine based on norm-linear-ConvNeXt-TCN

The supervisory control and data acquisition (SCADA) system of wind turbines continuously collects a large amount of monitoring data during their operation. These data contain abundant information about the operating status of the turbine components. Utilizing this information makes it feasible to provide early warnings and predict the health status of the wind turbine. However, due to the strong coupling between the various components of the wind turbine, the data exhibits complex spatiotemporal relationships, multiple state parameters, strong non-linearity, and noise interference, which brings great difficulty to anomaly detection of the wind turbine. This paper proposes a new method for detecting abnormal operating conditions of wind turbines, based on a cleverly designed multi-layer linear residual module and the improved temporal convolutional network (TCN) with a new norm-linear-ConvNeXt architecture (NLC-TCN). Initially, the NLC-TCN deep learning reconstruction model is trained with historical data of normal behavior to extract the spatiotemporal features of state parameters under normal operational conditions. Subsequently, the condition score of the unit is determined by calculating the average normalized root mean square error between the reconstructed data and actual data. The streaming peaks-over-threshold real-time calculation of the anomaly warning threshold, based on extreme value theory, is then used for preliminary fault monitoring. Moreover, by shielding the fault alarm for low wind speeds and implementing a continuous delay perception mechanism, issues related to wind speed fluctuations and internal and external interference are addressed, enabling early warning for faulty units. Finally, the effectiveness and reliability of the proposed method are validated through comparative experiments using actual offshore wind farm SCADA data. The performance of the proposed method surpasses that of other compared methods. Additionally, the results of the proposed method were evaluated using the uniform manifold approximation and projection dimensionality reduction technique and kernel density estimation.

Sodic-calcic alteration and transpressional shear along the Atacama fault system during IOCG mineralization, Copiapó, Chile

The Punta del Cobre district near Copiapó is a center of iron oxide-copper–gold (IOCG) mineralization spatially and temporally associated with regional sodic-calcic hydrothermal alteration, the Atacama fault system (AFS), and two phases of Early Cretaceous magmatism. Here, we investigate the spatiotemporal and geochemical relationships between magmatism, ductile deformation, and hydrothermal alteration along the ~ 200 to 300-m-thick steeply NW-dipping Sierra Chicharra shear zone, interpreted to be the major strand of the AFS. Mylonitic fabrics and oblique sinistral-reverse kinematic indicators together record coaxial flattening in a transpressional regime. Deformation on the AFS took place before, during, and after intrusion of the synkinematic Sierra Chicharra quartz diorite of the Coastal Cordillera arc at ~ 122 Ma and terminated before intrusion of the unstrained ~ 114 Ma Sierra Atacama diorite of the Copiapó batholith. Geochemical data show that the Copiapó batholith was more mafic and more K-rich than the calc-alkaline Coastal Cordillera arc. This time period thus overlaps IOCG mineralization in the Punta del Cobre district (~ 120 to 110 Ma). Multiple phases of sodic-calcic alteration in and around the AFS shear zone are recognized. Textures of altered rock in the shear zone show both synkinematic assemblages and post-kinematic hydrothermal oligoclase. A ~ 775-m-long andradite vein that cuts the shear zone formed broadly at the end of magmatism in the district (~ 95 Ma). Oxygen isotope ratios from the vein indicate that hydrothermal fluids were likely magmatically derived. Together, this work shows the AFS-related shear zone and nearby IOCG mineralization developed in a regional transpressional regime produced by SE-directed oblique convergence across a NE-striking shear zone. IOCG-related magmatic-hydrothermal fluids exploited this transcrustal shear zone to produce multiple episodes of regional sodic-calcic alteration formed from fluids exsolved from magmas or driven by the heat of the Coastal Cordillera arc and Copiapó batholith.

Simplified neural architecture for efficient human motion prediction in human-robot interaction

Human motion prediction is essential for safe and effective human-robot interaction, but modeling the intricate spatio-temporal dynamics inherent in human movement remains challenging. While recent methods have advanced motion prediction accuracy, they rely on complex neural architectures such as Graph Convolutional Networks and Recurrent Neural Networks, demanding extensive hyperparameter tuning. To overcome this limitation, we propose a streamlined multi-stage layer incremental perceptron (MSLP) architecture that achieves competitive results with far fewer parameters, improving efficiency. The MSLP eschews the complexity of prevalent networks and instead takes a stepped refinement approach to predict intricate motions using a lightweight model. This simplified yet effective design enables nuanced learning of spatio-temporal relationships without exhaustive tuning. Specifically, the MSLP incorporates two key components: a multi-channel feature extraction and enhancement block (MFE-block) and a temporal feature extraction module (TFE-block). The MFE-block strengthens the representation of each action node by integrating multi-dimensional action features. The TFE-block then captures the contextual relationships between actions over time. Together, the MFE-block and TFE-block allow the MSLP to model the complex spatial and temporal dynamics of human movement using a streamlined architecture and minimal parameter tuning. When evaluated on established datasets including Human3.6 M, CMU, 3DPW, and AMASS, the proposed MSLP method achieves improved accuracy gains of 16.7%-67.0% over existing state-of-the-art techniques. Additionally, the MSLP significantly reduces the number of parameters by 35.7%-99.5% compared to prior architectures. We find that the proposed MSLP significantly improves both long-term prediction and generalization capabilities of the model.