Spatio-temporal Visualization Research Articles

Intracellular dynamics of ubiquitin-like 3 visualized using an inducible fluorescent timer expression system.

Exosomes are small extracellular vesicles (sEVs) secreted via multivesicular bodies (MVBs)/late endosomes and mediators of cell-cell communication. We previously reported a novel post-translational modification by ubiquitin-like 3 (UBL3). UBL3 is localized in MVBs and the plasma membrane and released outside as sEVs, including exosomes. Approximately 60% of proteins sorted in sEVs are affected by UBL3 and localized in various organelles, the plasma membrane, and the cytosol, suggesting that its dynamic movement in the cell before entering the MVBs. To examine the intracellular dynamics of UBL3, we constructed a sophisticated visualization system via fusing fluorescent timers that changed from blue to red form over time with UBL3 and by its expression under Tet-on regulation. Intriguingly, we found that after synthesis, UBL3 was initially distributed within the cytosol. Subsequently, UBL3 was localized to MVBs and the plasma membrane and finally showed predominant accumulation in MVBs. Furthermore, by super-resolution microscopy analysis, UBL3 was found to be associated with one of its substrates, α-tubulin, in the cytosol, and the complex was subsequently transported to MVBs. This spatiotemporal visualization system for UBL3 will form a basis for further studies to elucidate when and where UBL3 associates with its substrates/binding proteins before localization in MVBs.

Dynamics‐Oriented Underwater Mechanoreception Interface for Simultaneous Flow and Contact Perception

The lack of a sufficient and efficient way to simultaneously perceive general underwater mechanical stimuli, physical contact, and fluidic flow has been a bottleneck for many aquatic applications. To address this challenge, dynamics‐oriented underwater mechanoreceptor interface (DOUMI), a bioinspired mechanoreception system that realizes simultaneous contact and flow perception using a single receptor, is introduced. This receptor, response‐elevated‐and‐expanded hair‐like tactile mechanoreceptor (REEM), is inspired by the mechanoreceptive mechanism of aquatic arthropods. REEM combines structural features from different mechanoreceptive sensilla, enabling it to capture a wide range of stimulus dynamics. Under different stimuli, REEM encodes stimuli dynamics as its oscillations with distinct spectral attributes. Those oscillations are efficiently transferred through mechanical processes and imaging, enabling vision‐based extraction and further analysis. Therefore, by evaluating the oscillation dynamics with tailored wavelet‐based indices, DOUMI can distinguish between contact‐ and flow‐induced oscillations at each receptor unit with 90.5% accuracy. Furthermore, DOUMI provides comprehensive 2D mechanoreception with a scalable array of REEMs, delivering capabilities like stimuli spatiotemporal visualization, flow trend detection, and scenario classification with an accuracy of 99.5%. With its robustness and operational efficiency in underwater environments, DOUMI can be easily adapted to existing applications using common materials and hardware, establishing a new, streamlined paradigm for underwater general mechanoreception.

GMMID: genetically modified mice information database.

Genetically engineered mouse models (GEMMs) are vital for elucidating gene function and disease mechanisms. An overwhelming number of GEMM lines have been generated, but endeavors to collect and organize the information of these GEMMs are seriously lagging behind. Only a few databases are developed for the information of current GEMMs, and these databases lack biological descriptions of allele compositions, which poses a challenge for nonexperts in mouse genetics to interpret the genetic information of these mice. Moreover, these databases usually do not provide information on human diseases related to the GEMM, which hinders the dissemination of the insights the GEMM provides as a human disease model. To address these issues, we developed an algorithm to annotate all the allele compositions that have been reported with Python programming and have developed the genetically modified mice information database (GMMID; http://www.gmmid.cn), a user-friendly database that integrates information on GEMMs and related diseases from various databases, including National Center for Biotechnology Information, Mouse Genome Informatics, Online Mendelian Inheritance in Man, International Mouse Phenotyping Consortium, and Jax lab. GMMID provides comprehensive genetic information on >70 055 alleles, 65 520 allele compositions, and ∼4000 diseases, along with biologically meaningful descriptions of alleles and allele combinations. Furthermore, it provides spatiotemporal visualization of anatomical tissues mentioned in these descriptions, shown alongside the allele compositions. Compared to existing mouse databases, GMMID considers the needs of researchers across different disciplines and presents obscure genetic information in an intuitive and easy-to-understand format. It facilitates users in obtaining complete genetic information more efficiently, making it an essential resource for cross-disciplinary researchers. Database URL: http://www.gmmid.cn.

A user study of visualisations of spatio-temporal eye tracking data

AbstractEye movements have a spatial (where people look), but also a temporal (when people look) component. Various types of visualizations have been proposed that take this spatio-temporal nature of the data into account, but it is unclear how well each one can be interpreted and whether such interpretation depends on the question asked about the data or the nature of the dataset that is being visualised. In this study, four spatio-temporal visualization techniques for eye movements (chord diagram, scan path, scarf plot, space-time cube) were compared in a user study. Participants $$(N = 25)$$ ( N = 25 ) answered three questions (what region first, what region most, which regions most between) about each visualization, which was based on two types of datasets (eye movements towards adverts, eye movements towards pairs of gambles). Accuracy of the answers depended on a combination of the dataset, the question that needed to answered, and the type of visualization. For most questions, the scan path, which did not use area of interest (AOI) information, resulted in lower accuracy than the other graphs. This suggests that AOIs improve the information conveyed by graphs. No effects of experience with reading graphs (for work or not for work) or education on accuracy of the answer was found. The results therefore suggest that there is no single best visualisation of the spatio-temporal aspects of eye movements. When visualising eye movement data, a user study may therefore be beneficial to determine the optimal visualization of the dataset and research question at hand. Graphical abstract

Laser-Induced Stress-Driven Nanoplate Jumping Visualized by Ultrafast Electron Microscopy.

Understanding laser-induced jumping has attracted great interest in nanomaterial launching and transfer but requires a high spatiotemporal resolution visualization. Here, we report a jumping dynamics of nanoplate driven by ultrafast laser-induced stress using time-resolved transmission electron microscopy. Single-shot imaging reveals a nondestructive launching of gold nanoplates in several nanoseconds after the pulsed femtosecond laser excitation. The temperature rise and acoustic vibration, derived from ultrafast electron crystallography with a picosecond time resolution, confirm the existence of a laser-induced elastic stress wave. The generation, propagation, and reflection of thermal stress waves are further clarified by atomic simulation. The nonequilibrium ultrafast laser heating produces a compressive stress wave within several picoseconds, constrained by the supporting substrate under nanoplate to provide thrust force. This compressive stress is subsequently reflected into tensile stress by the substrate, promoting the nanoplate to jump off the substrate. Furthermore, the uneven interface adhesion results in the jumping flip of nanoplates, as well as, diminished their jumping speed. This study unveils the jumping regime driven by impulsive laser-excited stress and offers understanding of light-matter interaction.

LinG3D: visualizing the spatio-temporal dynamics of clonal evolution.

Cancers are spatially heterogenous, thus their clonal evolution, especially following anti-cancer treatments, depends on where the mutated cells are located within the tumor tissue. For example, cells exposed to different concentrations of drugs, such as cells located near the vessels in contrast to those residing far from the vasculature, can undergo a different evolutionary path. However, classical representations of cell lineage trees do not account for this spatial component of emerging cancer clones. Here, we propose routines to trace spatial and temporal clonal evolution in computer simulations of the tumor evolution models. The LinG3D (Lineage Graphs in 3D) is an open-source collection of routines (in MATLAB, Python, and R) that enables spatio-temporal visualization of clonal evolution in a two-dimensional tumor slice from computer simulations of the tumor evolution models. These routines draw traces of tumor clones in both time and space, and may include a projection of a selected microenvironmental factor, such as the drug or oxygen distribution within the tumor, if such a microenvironmental factor is used in the tumor evolution model. The utility of LinG3D has been demonstrated through examples of simulated tumors with different number of clones and, additionally, in experimental colony growth assay. This routine package extends the classical lineage trees, that show cellular clone relationships in time, by adding the space component to show the locations of cellular clones within the 2D tumor tissue patch from computer simulations of tumor evolution models.

A new XR-based human‐robot collaboration assembly system based on industrial metaverse

The industrial metaverse is an emerging paradigm of integrated information technologies. A human-robot collaboration assembly system is developed using extended reality (XR) and blockchain technologies. This system includes the physical, virtual reality (VR), augmented reality (AR) supporting, and blockchain systems. Virtual avatars represent physical operators and mobile robots in the XR environment. The VR system emulates real-time physical operations, facilitating pre-planning, online task optimization, and converting strategic decisions into executable instructions for physical robots. The AR system enhances the cognitive experience of intelligent operators, contributing to decision-making and VR system control through multi-modal interactions. The integration of the VR and AR systems in perception-decision-control enables real-time process optimization and the reverse control of physical operations. The system is implemented using a decentralized blockchain and end-edge-cloud collaborative network architecture to regulate interaction permissions for each scene characters, ensuring systematic functionality, robust communication, and comprehensive data management. A case study of a gearbox assembly is used to demonstrate the system’s development and efficacy. The results indicate the VR system's comprehensive perception and the AR system's detailed insights, yielding reliable information on global and local levels. The integration of artificial intelligence (AI) algorithms and spatiotemporal visualization during VR decision-making provides computational advantages. Incorporating human input into the AR system enables human-in-the-loop operational decision-making, providing optimal instructions to improve human-robot collaboration performance. Notably, the fusion of blockchain and end-edge-cloud architectures augments inter-system data exchange by efficiently managing and processing human-robot interactions and their diverse avatars. This system offers an innovative approach to human-robot collaboration assembly in manufacturing, with implications for analogous contexts.

Spatiotemporal visualization of a surface acoustic wave coupled to magnons across a submillimeter-long sample by pulsed laser interferometry

Spatiotemporal visualization of a surface acoustic wave coupled to magnons across a submillimeter-long sample by pulsed laser interferometry

Universal Approach to Direct Spatiotemporal Dynamic In Situ Optical Visualization of On-Catalyst Water Splitting Electrochemical Processes.

Electrochemical reactions are the unrivaled backbone of next-generation energy storage, energy conversion, and healthcare devices. However, the real-time visualization of electrochemical reactions remains the bottleneck for fully exploiting their intrinsic potential. Herein, for the first time, a universal approach to direct spatiotemporal-dynamic in situ optical visualization of pH-based as well as specific byproduct-based electrochemical reactions is performed. As a highly relevant and impactful example, in-operando optical visualization of on-catalyst water splitting processes is performed in neutral water/seawater. HPTS (8-hydroxypyrene-1,3,6-trisulfonicacid), known for its exceptional optical capability of detecting even the tiniest pH changes allows the unprecedented "spatiotemporal" real-time visualization at the electrodes. As a result, it is unprecedentedly revealed that at a critical cathode-to-anode distance, the bulk-electrolyte "self-neutralization" phenomenon can be achieved during the water splitting process, leading to the practical realization of enhanced additive-free neutral water splitting. Furthermore, it is experimentally unveiled that at increasing electrolyte flow rates, a swift and severe inhibition of the concomitantly forming acidic and basic 'fronts', developed at anode and cathode compartments are observed, thus acting as a "buffering" mechanism. To demonstrate the universal applicability of this elegant strategy which is not limited to pH changes, the technique is extended to visualization of hypochlorite/ chlorine at the anode during electrolysis of sea water using N-(4-butanoic acid) dansylsulfonamide (BADS). Thus, a unique experimental tool that allows real-time spatiotemporal visualization and simultaneous mechanistic investigation of complex electrochemical processes is developed that can be universally extended to various fields of research.

Spatio-temporal void fraction visualization in air-water two-phase flow regime transitions by combination of convolutional neural network and long short-term memory implemented into multiple current-voltage (MCV-CNN_LSTM)

Spatio-temporal void fraction in air-water two-phase flow regime transitions has been visualized by combination of convolutional neural network and long short-term memory implemented into multiple -current-voltage (MCV-CNN_LSTM). The MCV-ML is composed of two components, which are MCV-CNN_LSTM training and evaluation. In the first component, four steps are carried out as 1) true void fraction α measurement was conducted by wire-mesh sensor (WMS) as objective variables, 2) simulated MCV voltage U calculation by simulation of MCV measurement for explanatory variable, 3) CNN_LSTM training instance generation, and 4) CNN_LSTM model training and testing. In the second component, two additional steps are experimentally conducted which are 5) actual MCV voltage V measurement by MCV for input of trained ML, and 6) spatio-temporal void fraction α prediction. For the dataset generation, the experiment was conducted on air-water two phase flow regime transitions in vertical pipe with an inner diameter of 25 mm and a length of 350 mm from the elbow. The superficial velocity of liquid and gas phase are varied to present the flow regime transition of bubbly to slug flow. As a result, the estimated spatio-temporal void fraction αˆ by MCV-CNN_LSTM enables the visualization of detailed gas distribution in vertical upward two-phase flow. The spatial averaged instantaneous void fraction measured by WMS and estimated by MCV-CNN_LSTM show qualitative agreement in normalized cross correlation (INCC) of 0.364 on its temporal variation.

DNA Double Strand Break and Response Fluorescent Assays: Choices and Interpretation.

Accurately characterizing DNA double-stranded breaks (DSBs) and understanding the DNA damage response (DDR) is crucial for assessing cellular genotoxicity, maintaining genomic integrity, and advancing gene editing technologies. Immunofluorescence-based techniques have proven to be invaluable for quantifying and visualizing DSB repair, providing valuable insights into cellular repair processes. However, the selection of appropriate markers for analysis can be challenging due to the intricate nature of DSB repair mechanisms, often leading to ambiguous interpretations. This comprehensively summarizes the significance of immunofluorescence-based techniques, with their capacity for spatiotemporal visualization, in elucidating complex DDR processes. By evaluating the strengths and limitations of different markers, we identify where they are most relevant chronologically from DSB detection to repair, better contextualizing what each assay represents at a molecular level. This is valuable for identifying biases associated with each assay and facilitates accurate data interpretation. This review aims to improve the precision of DSB quantification, deepen the understanding of DDR processes, assay biases, and pathway choices, and provide practical guidance on marker selection. Each assay offers a unique perspective of the underlying processes, underscoring the need to select markers that are best suited to specific research objectives.

In-situ registration subtraction image segmentation algorithm for spatiotemporal visualization of copper adsorption onto corn stalk-derived pellet biochar by micro-computed tomography

The non-homogeneous structure and high-density ash composition of biochar matrix pose significant challenges in characterizing the dynamic changes of heavy metal adsorption onto biochar with micro-computed tomography (Micro-CT). A novel in-situ registration subtraction image segmentation method (IRS) was developed to enhance micro-CT characterization accuracy. The kinetics of Cu(II) adsorption onto pellet biochar derived from corn stalks were tested. Respectively, the IRS and traditional K-means algorithms were used for image segmentation to the in-situ three-dimensional (3D) visual characterization of the Cu(II) adsorption onto biochar. The results indicated that the IRS algorithm reduced interference from high-density biochar composition, and thus achieved more precise results (R2 = 0.95) than that of K-means (R2 = 0.72). The visualized dynamic migration of Cu(II) from surface adsorption to intraparticle diffusion reflexed the complex mechanism of heavy metal adsorption. The developed Micro-CT method with high generalizability has great potential for studying the process and mechanism of biochar heavy metal adsorption.

A Planning Support System for Monitoring Aging Neighborhoods in Germany

Many single-family homes built in Germany in the first decades following the Second World War are now occupied by elderly residents. If local conditions are unfavorable, a large number of these buildings may enter the real estate market in a short period of time and put pressure on the local housing market. Planners and decision-makers therefore need detailed spatiotemporal information about these neighborhoods to effectively address and counteract such developments. We present the design and implementation of a planning support system that can generate the required information. The architecture of this newly developed software consists of a composite, multitier framework to perform the complex tasks of data importation, data processing, and visualization. Legally mandated municipal population registers provide the key data for the calculation of indicators as a base for spatiotemporal analyses and visualizations. These registers offer high data quality in terms of completeness, logical consistency, spatial, and temporal and thematic accuracy. We demonstrate the implemented method using population data from a local government in a rural area in southwestern Germany. The results show that the new tool, which relies on open software components, is capable to identify and prioritize areas with particularly high levels of problem pressure. The tool can be used not only for analyses in a local context, but also at a regional level.

An Indigenous Computational Platform for Nowcasting and Forecasting Non-Linear Spread of COVID-19 across the Indian Sub-continent: A Geo-Temporal Visualization of Data

The rapid spread of the COVID-19 pandemic necessitated unprecedented collective action against coronavirus disease. In this light,we are proposing a novel online platform for the visualization of epidemiological data incorporating social determinants for understanding the patterns associated with the spread of COVID-19. The current AI computational platform combines modeling methodologies along with temporal geospatial visualization of COVID-19 data, providing real-time sharing of graphic analytical simulation of vulnerable hotspots of recurrent (nowcasting) and emergent (forecasting) infections visualized on a spatiotemporal scale on geoportals. The proposed study will be a secondary data analysis of primary data accessed from the national portal (Indian Council of Medical Research (ICMR)) incorporating 766 districts in India. Epidemiological data related to spatiotemporal visualization of the demographic spread of COVID-19 will be displayed using a compartmental socio-epidemiological model, reproduction number R, epi-curve diagrams as well as choropleth maps for different levels of administrative and development units at the district levels.

Genetically Encoded Fluorescence Resonance Energy Transfer Biosensor for Live-Cell Visualization of Lamin A Phosphorylation at Serine 22.

Extensive phosphorylation at serine 22 (pSer22) on lamin A is the hallmark of cell mitosis, which contributes to the breakdown of nuclear envelope. In the interphase, pSer22 lamin A exists in low abundance and is involved in mechanotransduction, virus infection, and gene expression. Numerous evidences emerge to support lamin A regulation on cell function and fate by phosphorylation. However, live-cell imaging tools for visualizing the dynamics of pSer22 lamin A are yet to be established. Herein, we developed a novel lamin A phosphorylation sensor (LAPS) based on fluorescence resonance energy transfer (FRET) with high sensitivity and specificity. We observed the dynamic lamin A phosphorylation during the cell cycle progression in single living cells: the increase of pSer22 modification when cells entered the mitosis and recovered upon the mitosis exit. Our biosensor also showed the gradual reduction of pSer22 modification during cell adhesion and in response to hypotonic environment. By applying LAPS, we captured the propagation of pSer22 modification from inside to outside of the inner nuclear membrane, which further led to the breakdown of nuclear envelope. Meanwhile, we found the synchronous phosphorylation of pSer22 lamin A and H3S10ph at mitosis entry. Inhibition of Aurora B, the responsible kinase for H3S10ph, could shorten the mitotic period without obvious effect on the pSer22 modification level of lamin A. Thus, LAPS allows the spatiotemporal visualization of the lamin A pSer22, which will be useful for elucidating the molecular mechanisms underlying cell mitosis and mechanoresponsive processes.

Interpretable spatio-temporal modeling for soil temperature prediction

Soil temperature (ST) is a crucial parameter in Earth system science. Accurate ST predictions provide invaluable insights; however, the “black box” nature of many deep learning approaches limits their interpretability. In this study, we present the Encoder-Decoder Model with Interpretable Spatio-Temporal Component (ISDNM) to enhance both ST prediction accuracy and its spatio-temporal interpretability. The ISDNM combines a CNN-encoder-decoder and an LSTM-encoder-decoder to improve spatio-temporal feature representation. It further uses linear regression and Uniform Manifold Approximation and Projection (UMAP) techniques for clearer spatio-temporal visualization of ST. The results show that the ISDNM model had the highest R2 ranging from 0.886 to 0.963 and the lowest RMSE ranging from 6.086 m3/m3 to 12.533 m3/m3 for different climate regions, and demonstrated superior performance than all the other DL models like CNN, LSTM, ConvLSTM models. The predictable component highlighted the remarkable similarity between Medium fine and Very fine soils in China. Additional, May and November emerged as crucial months, acting as inflection points in the annual ST cycle, shaping ISDNM model’s prediction capabilities.

Deep learning from atrioventricular plane displacement in patients with Takotsubo syndrome: lighting up the black-box.

The spatiotemporal deep convolutional neural network (DCNN) helps reduce echocardiographic readers' erroneous 'judgement calls' on Takotsubo syndrome (TTS). The aim of this study was to improve the interpretability of the spatiotemporal DCNN to discover latent imaging features associated with causative TTS pathophysiology. We applied gradient-weighted class activation mapping analysis to visualize an established spatiotemporal DCNN based on the echocardiographic videos to differentiate TTS (150 patients) from anterior wall ST-segment elevation myocardial infarction (STEMI, 150 patients). Forty-eight human expert readers interpreted the same echocardiographic videos and prioritized the regions of interest on myocardium for the differentiation. Based on visualization results, we completed optical flow measurement, myocardial strain, and Doppler/tissue Doppler echocardiography studies to investigate regional myocardial temporal dynamics and diastology. While human readers' visualization predominantly focused on the apex of the heart in TTS patients, the DCNN temporal arm's saliency visualization was attentive on the base of the heart, particularly at the atrioventricular (AV) plane. Compared with STEMI patients, TTS patients consistently showed weaker peak longitudinal displacement (in pixels) in the basal inferoseptal (systolic: 2.15 ± 1.41 vs. 3.10 ± 1.66, P < 0.001; diastolic: 2.36 ± 1.71 vs. 2.97 ± 1.69, P = 0.004) and basal anterolateral (systolic: 2.70 ± 1.96 vs. 3.44 ± 2.13, P = 0.003; diastolic: 2.73 ± 1.70 vs. 3.45 ± 2.20, P = 0.002) segments, and worse longitudinal myocardial strain in the basal inferoseptal (-8.5 ± 3.8% vs. -9.9 ± 4.1%, P = 0.013) and basal anterolateral (-8.6 ± 4.2% vs. -10.4 ± 4.1%, P = 0.006) segments. Meanwhile, TTS patients showed worse diastolic mechanics than STEMI patients (E'/septal: 5.1 ± 1.2 cm/s vs. 6.3 ± 1.5 cm/s, P < 0.001; S'/septal: 5.8 ± 1.3 cm/s vs. 6.8 ± 1.4 cm/s, P < 0.001; E'/lateral: 6.0 ± 1.4 cm/s vs. 7.9 ± 1.6 cm/s, P < 0.001; S'/lateral: 6.3 ± 1.4 cm/s vs. 7.3 ± 1.5 cm/s, P < 0.001; E/E': 15.5 ± 5.6 vs. 12.5 ± 3.5, P < 0.001). The spatiotemporal DCNN saliency visualization helps identify the pattern of myocardial temporal dynamics and navigates the quantification of regional myocardial mechanics. Reduced AV plane displacement in TTS patients likely correlates with impaired diastolic mechanics.

Actionable descriptors of spatiotemporal urban dynamics from large-scale mobile data: A case study in Lisbon city

Mobile phones share location records, offering the opportunity to monitor and understand emerging population dynamics in urban centers. With the aim of supporting urban planning, this study introduces a scalable methodology grounded on extracting and organizing spatiotemporal statistics from decomposed population density data. The proposed methodology serves three major purposes: (i) assess the predictability of spatiotemporal citizen density patterns; (ii) detect emerging spatiotemporal trends in population density; and (iii) uncover multi-level seasonality patterns with guarantees of actionability. Additionally, it makes available an open-access tool for deploying the proposed methodology and analyzing mobile phone network data with easy-to-use spatiotemporal visualization and navigation facilities. The results obtained from real-world, large-scale mobile data in Lisbon, Portugal, demonstrate the effectiveness and validity of the proposed methodology in extracting actionable statistics in linear time to guide both tactic and strategic urban planning.

Spatio-Temporal variability of soil primary macro and micro nutrients – An empirical analysis on Anaimalai block, Coimbatore District, India

This study reveals Spatio-temporal prognostication of macro and micronutrients present in entisols, inceptisols and alfisols soil groups at Anaimalai block of Coimbatore district, India through empirical analysis. The soil primary macronutrients and micronutrients for the period 2021–2032 were procured from Autoregressive Integrated Moving Average (ARIMA), Linear Regression (LR) and Logistic Regression (Log R) models by setting out 2005–2020 soil nutrient datasets as input. 1627 soil samples were taken out through composite sampling method in the month of February 2021. The laboratory soil analysis was carried out for macro and micronutrients computation. Computed values were validated with the 2021 soil nutrient empirical models. R2 values 0.88, 0.83, 0.85, 0.89, 0.88, 0.92, 0.90, 0.92 and 0.85 for soil pH, EC, Available nitrogen, Available phosphorus, Available potassium, Iron, Manganese, Copper and Zinc respectively divulges the reliability of ARIMA model in forecasting the soil nutrients and are spatially exposed. Spatio-temporal visualization of soil nutrients for a decade with large nutrient database as input along with field validation brings out promising forecasting model that reverberates on futuristic policy making for attaining sustainable agro-productivity.

Interpretable hierarchical error correction GRU model for effective observation selection

Interpretability of extended target effective observation selection model in artificial intelligence (AI) symbolizes a severe challenge: both accuracy effective observation selection while ensuring real-time performance with an extended target multiple observation model and physical interpretability are vital. To meet the challenge, this paper proposes a novel hierarchical Gated Recurrent Unit (GRU) interpretable model. The contribution of this study lies in filling the research gap in the field of effective observation processing for extended targets and intuitively explaining the spatiotemporal visualization of the model. Compared with other interpretable models, the proposed model has improved data clustering and feature extraction capabilities, noise robustness, and computational resource consumption in the face of complex extended target multivariate physical factors. More importantly, interpretable studies are presented on the spatiotemporal characteristics of the multiple observation model of the extended target. The experimental results in real-world scenarios demonstrate that the proposed model is better than benchmark models in terms of prediction performance. The visualization and interpretation of the integrated predictive model reflects the reasonability of the proposed ensemble model.