Щоб переглянути інші типи публікацій з цієї теми, перейдіть за посиланням: Spatial granularities.
Статті в журналах з теми "Spatial granularities"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-27 статей у журналах для дослідження на тему "Spatial granularities".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте статті в журналах для різних дисциплін та оформлюйте правильно вашу бібліографію.
1
Tanaka, Yusuke, Tomoharu Iwata, Toshiyuki Tanaka, Takeshi Kurashima, Maya Okawa, and Hiroyuki Toda. "Refining Coarse-Grained Spatial Data Using Auxiliary Spatial Data Sets with Various Granularities." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 5091–99. http://dx.doi.org/10.1609/aaai.v33i01.33015091.
Повний текст джерелаАнотація:
We propose a probabilistic model for refining coarse-grained spatial data by utilizing auxiliary spatial data sets. Existing methods require that the spatial granularities of the auxiliary data sets are the same as the desired granularity of target data. The proposed model can effectively make use of auxiliary data sets with various granularities by hierarchically incorporating Gaussian processes. With the proposed model, a distribution for each auxiliary data set on the continuous space is modeled using a Gaussian process, where the representation of uncertainty considers the levels of granularity. The finegrained target data are modeled by another Gaussian process that considers both the spatial correlation and the auxiliary data sets with their uncertainty. We integrate the Gaussian process with a spatial aggregation process that transforms the fine-grained target data into the coarse-grained target data, by which we can infer the fine-grained target Gaussian process from the coarse-grained data. Our model is designed such that the inference of model parameters based on the exact marginal likelihood is possible, in which the variables of finegrained target and auxiliary data are analytically integrated out. Our experiments on real-world spatial data sets demonstrate the effectiveness of the proposed model.
2
Shook, Eric, and Shaowen Wang. "Investigating the Influence of Spatial and Temporal Granularities on Agent-Based Modeling." Geographical Analysis 47, no. 4 (July 20, 2015): 321–48. http://dx.doi.org/10.1111/gean.12080.
Повний текст джерела
3
Yu, WANG, LIAN Yuntao, FENG Qi, WANG Zhijun, ZHAO Weijun, LIU Juanjuan, LU Shiguo, and ZHANG Xinyu. "Effects of dam interception on the spatial distribution of sediment granularities in Heihe River." Journal of Lake Sciences 31, no. 5 (2019): 1459–67. http://dx.doi.org/10.18307/2019.0505.
Повний текст джерела
4
Wallgrün, Jan Oliver, Jinlong Yang, and Alexander Klippel. "Cognitive Evaluation of Spatial Formalisms." International Journal of Cognitive Informatics and Natural Intelligence 8, no. 1 (January 2014): 1–17. http://dx.doi.org/10.4018/ijcini.2014010101.
Повний текст джерелаАнотація:
The authors present four human behavioral experiments to address the question of intuitive granularities in fundamental spatial relations as they can be found in formal spatial calculi. These calculi focus on invariant characteristics under certain (especially topological) transformations. Of particular interest to this article is the concept of two spatially extended entities overlapping each other. The overlap concept has been extensively treated in Galton's mode of overlap calculus (Galton, 1998). In the first two experiments, the authors used a category construction task to calibrate this calculus against behavioral data and found that participants adopted a very coarse view on the concept of overlap and distinguished only between three general relations: proper part, overlap, and non-overlap. In the following two experiments, the authors changed the instructions to explicitly address the possibility that humans could be swayed to adopt a more detailed level of granularity, that is, the authors encouraged them to create as many meaningful groups as possible. The results show that the three relations identified in the first two experiments (overlap, non-overlap, and proper part) are very robust and a natural level of granularity across all four experiments. However, the results also reveal that contextual factors gain more influence at finer levels of granularity.
5
Mossakowski, Till, and Reinhard Moratz. "Relations Between Spatial Calculi About Directions and Orientations." Journal of Artificial Intelligence Research 54 (November 1, 2015): 277–308. http://dx.doi.org/10.1613/jair.4631.
Повний текст джерелаАнотація:
Qualitative spatial descriptions characterize essential properties of spatial objects or configurations by relying on relative comparisons rather than measuring. Typically, in qualitative approaches only relatively coarse distinctions between configurations are made. Qualitative spatial knowledge can be used to represent incomplete and underdetermined knowledge in a systematic way. This is especially useful if the task is to describe features of classes of configurations rather than individual configurations. Although reasoning with them is generally NP-hard, relative directions are important because they play a key role in human spatial descriptions and there are several approaches how to represent them using qualitative methods. In these approaches directions between spatial locations can be expressed as constraints over infinite domains, e.g. the Euclidean plane. The theory of relation algebras has been successfully applied to this field. Viewing relation algebras as universal algebras and applying and modifying standard tools from universal algebra in this work, we (re)define notions of qualitative constraint calculus, of homomorphism between calculi, and of quotient of calculi. Based on this method we derive important properties for spatial calculi from corresponding properties of related calculi. From a conceptual point of view these formal mappings between calculi are a means to translate between different granularities.
6
Qiao, Xuning, Liang Liu, Yongju Yang, Yangyang Gu, and Jinchan Zheng. "Urban Expansion Assessment Based on Optimal Granularity in the Huaihe River Basin of China." Sustainability 14, no. 20 (October 17, 2022): 13382. http://dx.doi.org/10.3390/su142013382.
Повний текст джерелаАнотація:
Determining the optimal granularity, which has often been ignored in the analysis of urban expansion and its landscape pattern, is the core problem in landscape ecology research. Here, we calculate the optimal granularities for differently sized cities in the Huaihe River Basin of China based on scale transformation and area loss evaluation. Accordingly, we construct a landscape index and urban land density function to analyze urban expansion and landscape pattern. The results can be summarized as follows. (1) Within the first scale domain of the landscape indices, the optimal granularities of Zhengzhou, Xuzhou, Yancheng, Xinyang, and Bozhou are 60 m, 50 m, 40 m, 40 m, and 40 m, respectively, which are the optimal units in the study of urban expansion. (2) The urban land density decreases from the urban center to the outskirts, the urban core of each city is more compact than the outskirts, and the land density curve parameter α of Zhengzhou is the largest at 4.693 and its urban core the most compact. (3) There are significant spatial and temporal differences in the urban land densities of differently sized cities. The urban land density functions of different cities are similar before 2000; after that, they are similar to the standard inverse S-shaped function and the land use density curve of large cities is closer to the standard inverse S-shaped function than that of small- and medium-sized cities. (4) Large cities have faster expansion, much larger land density curve parameter c than medium- and small-cities, stronger linkage development with surrounding areas, and a higher degree of urban centralization. Urban expansion compactness was influenced by urban locations and functions except for urban sizes. This study offers a method for identifying the optimal granularities for differently sized cities and also provides information for the decision-making efforts that concern the rapid urbanization in major grain-producing areas of China.
7
EL-GERESY, BAHER A., and ALIA I. ABDELMOTY. "Topological representation and reasoning in space." Artificial Intelligence for Engineering Design, Analysis and Manufacturing 14, no. 5 (November 2000): 373–89. http://dx.doi.org/10.1017/s0890060400145032.
Повний текст джерелаАнотація:
In this article, an approach is presented for the representation and reasoning over qualitative spatial relations. A set-theoretic approach is used for representing the topology of objects and underlying space by retaining connectivity relationships between objects and space components in a structure, denoted, adjacency matrix. Spatial relations are represented by the intersection of components, and spatial reasoning is achieved by the application of general rules for the propagation of the intersection constraints between those components. The representation approach is general and can be adapted for different space resolutions and granularities of relations. The reasoning mechanism is simple and the spatial compositions are achieved in a finite definite number of steps, controlled by the complexity needed in the representation of objects and the granularity of the spatial relations required. The application of the method is presented over geometric structures that takes into account qualitative surface height information. It is also shown how directional relationships can be used in a hybrid approach for richer composition scenarios. The main advantage of this work is that it offers a unified platform for handling different relations in the qualitative space, which is a step toward developing general spatial reasoning engines for large spatial databases.
8
Jian, Yang, Jinhong Li, Lu Wei, Lei Gao, and Fuqi Mao. "Spatiotemporal DeepWalk Gated Recurrent Neural Network: A Deep Learning Framework for Traffic Learning and Forecasting." Journal of Advanced Transportation 2022 (April 18, 2022): 1–11. http://dx.doi.org/10.1155/2022/4260244.
Повний текст джерелаАнотація:
As a typical spatiotemporal problem, there are three main challenges in traffic forecasting. First, the road network is a nonregular topology, and it is difficult to extract complex spatial dependence accurately. Second, there are short- and long-term dependencies between traffic dates. Third, there are many other factors besides the influence of spatiotemporal dependence, such as semantic characteristics. To address these issues, we propose a spatiotemporal DeepWalk gated recurrent unit model (ST-DWGRU), a deep learning framework that fuses spatial, temporal, and semantic features for traffic speed forecasting. In the framework, the spatial dependency between nodes of an entire road network is extracted by graph convolutional network (GCN), whereas the temporal dependency between speeds is captured by a gated recurrent unit network (GRU). DeepWalk is used to extract semantic information from road networks. Three publicly available datasets with different time granularities of 15, 30, and 60 min are used to validate the short- and long-time prediction effect of this model. The results show that the ST-DWGRU model significantly outperforms the state-of-the-art baselines.
9
Beard, Kate, Heather Deese, and Neal R. Pettigrew. "A Framework for Visualization and Exploration of Events." Information Visualization 7, no. 2 (December 20, 2007): 133–51. http://dx.doi.org/10.1057/palgrave.ivs.9500165.
Повний текст джерелаАнотація:
The expanding deployment of sensor systems that capture location, time, and multiple thematic variables is increasing the need for exploratory spatio-temporal data analysis tools. Geographic information systems (GIS) and time series analysis tools support exploration of spatial and temporal patterns respectively and independently, but tools for the exploration of both dimensions within a single system are relatively rare. The contribution of this research is a framework for the visualization and exploration of spatial, temporal, and thematic dimensions of sensor-based data. The unit of analysis is an event, a spatio-temporal data type extracted from sensor data. The conceptual framework suggests an approach for design layout that can be flexibly modified to explore spatial and temporal trends, temporal relationships among events, periodic temporal patterns, the timing of irregularly repeating events, event–event relationships in terms of thematic attributes, and event patterns at different spatial and temporal granularities. Flexible assignment of spatial, temporal, and thematic categories to a set of graphical interface elements that can be easily rearranged provides exploratory power as well as a generalizable design layout structure. The framework is illustrated with events extracted from Gulf of Maine Ocean Observing System data but the approach has broad application to other domains and applications in which time, space, and attributes need to be considered in conjunction.
10
Zhao, Zilin, Yuanying Chi, Zhiming Ding, Mengmeng Chang, and Zhi Cai. "Latent Semantic Sequence Coding Applied to Taxi Travel Time Estimation." ISPRS International Journal of Geo-Information 12, no. 2 (January 31, 2023): 44. http://dx.doi.org/10.3390/ijgi12020044.
Повний текст джерелаАнотація:
Taxi travel time estimation based on real-time traffic flow collection in IoT has been well explored; however, it becomes a challenge to use the limited taxi data to estimate the travel time. Most of the existing methods in this scenario rely on shallow feature engineering. Nevertheless, they have limited performance in learning complex moving patterns. Thus, a Latent Semantic Pulse Sequence-based Deep Neural Network (LSPS-DNN) is proposed in this paper to improve the taxi travel time estimation performance by constructing a latent semantic propagation graph representing the latent path sequence. It first extracts the shallow modal features of trips, such as the time period and spatial location at different granularities. The representation of the pulse propagation graph is then extracted from shallow spatial features using a Pulse Coupled Neural Network (PCNN). Further, the propagation graph is encoded with negative sampling to obtain the embedding of deep propagation features between ODs. Meanwhile, we conduct deep network learning based on the Chengdu and NYC taxi datasets; our experimental evaluation results show it has a better performance compared to traditional feature construction methods.
11
Zheng, K., D. Gu, F. Fang, Y. Wang, H. Liu, W. Zhao, M. Zhang, and Q. Li. "VISUALIZATION OF SPATIO-TEMPORAL RELATIONS IN MOVEMENT EVENT USING MULTI-VIEW." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-2/W7 (September 14, 2017): 1469–76. http://dx.doi.org/10.5194/isprs-archives-xlii-2-w7-1469-2017.
Повний текст джерелаАнотація:
Spatio-temporal relations among movement events extracted from temporally varying trajectory data can provide useful information about the evolution of individual or collective movers, as well as their interactions with their spatial and temporal contexts. However, the pure statistical tools commonly used by analysts pose many difficulties, due to the large number of attributes embedded in multi-scale and multi-semantic trajectory data. The need for models that operate at multiple scales to search for relations at different locations within time and space, as well as intuitively interpret what these relations mean, also presents challenges. Since analysts do not know where or when these relevant spatio-temporal relations might emerge, these models must compute statistical summaries of multiple attributes at different granularities. In this paper, we propose a multi-view approach to visualize the spatio-temporal relations among movement events. We describe a method for visualizing movement events and spatio-temporal relations that uses multiple displays. A visual interface is presented, and the user can interactively select or filter spatial and temporal extents to guide the knowledge discovery process. We also demonstrate how this approach can help analysts to derive and explain the spatio-temporal relations of movement events from taxi trajectory data.
12
Liu, Zhao-ge, Xiang-yang Li, and Xiao-han Zhu. "A full-view scenario model for urban waterlogging response in a big data environment." Open Geosciences 13, no. 1 (January 1, 2021): 1432–47. http://dx.doi.org/10.1515/geo-2020-0317.
Повний текст джерелаАнотація:
Abstract The emergence of big data is breaking the spatial and time limitations of urban waterlogging scenario description. The scenario data of different dimensions (e.g., administrative levels, sectors, granularities, and time) have become highly integrated. Accordingly, a structural and systematic model is needed to represent waterlogging scenarios for more efficient waterlogging response decision-making. In this article, a full-view urban waterlogging scenario is first defined and described from four dimensions. Next a structured representation of scenario element is given based on knowledge unit method. The full-view scenario model is then constructed by extracting the scenario correlation structures between different dimensions (called scenario nesting), i.e., inheritance nesting, feedback nesting, aggregation nesting, and selection nesting. Finally, a real-world case study in Wuhan East Lake High-tech Development Zone, China is evaluated to verify the reasonability of the full-view model. The results show that the proposed model effectively integrates scenario data from different dimensions, which helps generate the complete key scenario information for urban waterlogging decision-making. The full-view scenario model is expected to be applicable for other disasters under big data environment.
13
Yin, Quanjun, Long Qin, Xiaocheng Liu, and Yabing Zha. "Incremental Construction of Generalized Voronoi Diagrams on Pointerless Quadtrees." Mathematical Problems in Engineering 2014 (2014): 1–14. http://dx.doi.org/10.1155/2014/456739.
Повний текст джерелаАнотація:
In robotics, Generalized Voronoi Diagrams (GVDs) are widely used by mobile robots to represent the spatial topologies of their surrounding area. In this paper we consider the problem of constructing GVDs on discrete environments. Several algorithms that solve this problem exist in the literature, notably the Brushfire algorithm and its improved versions which possess local repair mechanism. However, when the area to be processed is very large or is of high resolution, the size of the metric matrices used by these algorithms to compute GVDs can be prohibitive. To address this issue, we propose an improvement on the current algorithms, using pointerless quadtrees in place of metric matrices to compute and maintain GVDs. Beyond the construction and reconstruction of a GVD, our algorithm further provides a method to approximate roadmaps in multiple granularities from the quadtree based GVD. Simulation tests in representative scenarios demonstrate that, compared with the current algorithms, our algorithm generally makes an order of magnitude improvement regarding memory cost when the area is larger than210×210. We also demonstrate the usefulness of the approximated roadmaps for coarse-to-fine pathfinding tasks.
14
Guo, Jinjin, Zhiguo Gong, and Longbing Cao. "dhCM: Dynamic and Hierarchical Event Categorization and Discovery for Social Media Stream." ACM Transactions on Intelligent Systems and Technology 12, no. 5 (October 31, 2021): 1–25. http://dx.doi.org/10.1145/3470888.
Повний текст джерелаАнотація:
The online event discovery in social media based documents is useful, such as for disaster recognition and intervention. However, the diverse events incrementally identified from social media streams remain accumulated, ad hoc, and unstructured. They cannot assist users in digesting the tremendous amount of information and finding their interested events. Further, most of the existing work is challenged by jointly identifying incremental events and dynamically organizing them in an adaptive hierarchy. To address these problems, this article proposes d ynamic and h ierarchical C ategorization M odeling (dhCM) for social media stream. Instead of manually dividing the timeframe, a multimodal event miner exploits a density estimation technique to continuously capture the temporal influence between documents and incrementally identify online events in textual, temporal, and spatial spaces. At the same time, an adaptive categorization hierarchy is formed to automatically organize the documents into proper categories at multiple levels of granularities. In a nonparametric manner, dhCM accommodates the increasing complexity of data streams with automatically growing the categorization hierarchy over adaptive growth. A sequential Monte Carlo algorithm is used for the online inference of the dhCM parameters. Extensive experiments show that dhCM outperforms the state-of-the-art models in terms of term coherence, category abstraction and specialization, hierarchical affinity, and event categorization and discovery accuracy.
15
Qing, Yuhao, Quanzhen Huang, Liuyan Feng, Yueyan Qi, and Wenyi Liu. "Multiscale Feature Fusion Network Incorporating 3D Self-Attention for Hyperspectral Image Classification." Remote Sensing 14, no. 3 (February 5, 2022): 742. http://dx.doi.org/10.3390/rs14030742.
Повний текст джерелаАнотація:
In recent years, the deep learning-based hyperspectral image (HSI) classification method has achieved great success, and the convolutional neural network (CNN) method has achieved good classification performance in the HSI classification task. However, the convolutional operation only works with local neighborhoods, and is effective in extracting local features. It is difficult to capture interactive features over long distances, which affects the accuracy of classification to some extent. At the same time, the data from HSI have the characteristics of three-dimensionality, redundancy, and noise. To solve these problems, we propose a 3D self-attention multiscale feature fusion network (3DSA-MFN) that integrates 3D multi-head self-attention. 3DSA-MFN first uses different sized convolution kernels to extract multiscale features, samples the different granularities of the feature map, and effectively fuses the spatial and spectral features of the feature map. Then, we propose an improved 3D multi-head self-attention mechanism that provides local feature details for the self-attention branch, and fully exploits the context of the input matrix. To verify the performance of the proposed method, we compare it with six current methods on three public datasets. The experimental results show that the proposed 3DSA-MFN achieves competitive classification and highlights the HSI classification task.
16
Bayoumi, Randa Mohamed, Elsayed E. Hemayed, Mohammad Ehab Ragab, and Magda B. Fayek. "Person Re-Identification via Pyramid Multipart Features and Multi-Attention Framework." Big Data and Cognitive Computing 6, no. 1 (February 9, 2022): 20. http://dx.doi.org/10.3390/bdcc6010020.
Повний текст джерелаАнотація:
Video-based person re-identification has become quite attractive due to its importance in many vision surveillance problems. It is a challenging topic due to the inter/intra changes, occlusion, and pose variations involved. In this paper, we propose a pyramid-attentive framework that relies on multi-part features and multiple attention to aggregate features of multi-levels and learns attention-based representations of persons through various aspects. Self-attention is used to strengthen the most discriminative features in the spatial and channel domains and hence capture robust global information. We propose the use of part-relation attention between different multi-granularities of features’ representation to focus on learning appropriate local features. Temporal attention is used to aggregate temporal features. We integrate the most robust features in the global and multi-level views to build an effective convolution neural network (CNN) model. The proposed model outperforms the previous state-of-the art models on three datasets. Notably, using the proposed model enables the achievement of 98.9% (a relative improvement of 2.7% on the GRL) top1 accuracy and 99.3% mAP on the PRID2011, and 92.8% (a relative improvement of 2.4% relative to GRL) top1 accuracy on iLIDS-vid. We also explore the generalization ability of our model on a cross dataset.
17
Ma, Xinyi, Zhifeng Xiao, Hong-sik Yun, and Seung-Jun Lee. "Semantic Multigranularity Feature Learning for High-Resolution Remote Sensing Image Scene Classification." Applied Sciences 11, no. 19 (October 3, 2021): 9204. http://dx.doi.org/10.3390/app11199204.
Повний текст джерелаАнотація:
High-resolution remote sensing image scene classification is a challenging visual task due to the large intravariance and small intervariance between the categories. To accurately recognize the scene categories, it is essential to learn discriminative features from both global and local critical regions. Recent efforts focus on how to encourage the network to learn multigranularity features with the destruction of the spatial information on the input image at different scales, which leads to meaningless edges that are harmful to training. In this study, we propose a novel method named Semantic Multigranularity Feature Learning Network (SMGFL-Net) for remote sensing image scene classification. The core idea is to learn both global and multigranularity local features from rearranged intermediate feature maps, thus, eliminating the meaningless edges. These features are then fused for the final prediction. Our proposed framework is compared with a collection of state-of-the-art (SOTA) methods on two fine-grained remote sensing image scene datasets, including the NWPU-RESISC45 and Aerial Image Datasets (AID). We justify several design choices, including the branch granularities, fusion strategies, pooling operations, and necessity of feature map rearrangement through a comparative study. Moreover, the overall performance results show that SMGFL-Net consistently outperforms other peer methods in classification accuracy, and the superiority is more apparent with less training data, demonstrating the efficacy of feature learning of our approach.
18
Elsaka, Tarek, Imad Afyouni, Ibrahim Hashem, and Zaher Al Aghbari. "Spatio-Temporal Sentiment Mining of COVID-19 Arabic Social Media." ISPRS International Journal of Geo-Information 11, no. 9 (September 2, 2022): 476. http://dx.doi.org/10.3390/ijgi11090476.
Повний текст джерелаАнотація:
Since the recent outbreak of COVID-19, many scientists have started working on distinct challenges related to mining the available large datasets from social media as an effective asset to understand people’s responses to the pandemic. This study presents a comprehensive social data mining approach to provide in-depth insights related to the COVID-19 pandemic and applied to the Arabic language. We first developed a technique to infer geospatial information from non-geotagged Arabic tweets. Secondly, a sentiment analysis mechanism at various levels of spatial granularities and separate topic scales is introduced. We applied sentiment-based classifications at various location resolutions (regions/countries) and separate topic abstraction levels (subtopics and main topics). In addition, a correlation-based analysis of Arabic tweets and the official health providers’ data will be presented. Moreover, we implemented several mechanisms of topic-based analysis using occurrence-based and statistical correlation approaches. Finally, we conducted a set of experiments and visualized our results based on a combined geo-social dataset, official health records, and lockdown data worldwide. Our results show that the total percentage of location-enabled tweets has increased from 2% to 46% (about 2.5M tweets). A positive correlation between top topics (lockdown and vaccine) and the COVID-19 new cases has also been recorded, while negative feelings of Arab Twitter users were generally raised during this pandemic, on topics related to lockdown, closure, and law enforcement.
19
Huang, Lan, Jia Zeng, Shiqi Sun, Wencong Wang, Yan Wang, and Kangping Wang. "Coarse-Grained Pruning of Neural Network Models Based on Blocky Sparse Structure." Entropy 23, no. 8 (August 13, 2021): 1042. http://dx.doi.org/10.3390/e23081042.
Повний текст джерелаАнотація:
Deep neural networks may achieve excellent performance in many research fields. However, many deep neural network models are over-parameterized. The computation of weight matrices often consumes a lot of time, which requires plenty of computing resources. In order to solve these problems, a novel block-based division method and a special coarse-grained block pruning strategy are proposed in this paper to simplify and compress the fully connected structure, and the pruned weight matrices with a blocky structure are then stored in the format of Block Sparse Row (BSR) to accelerate the calculation of the weight matrices. First, the weight matrices are divided into square sub-blocks based on spatial aggregation. Second, a coarse-grained block pruning procedure is utilized to scale down the model parameters. Finally, the BSR storage format, which is much more friendly to block sparse matrix storage and computation, is employed to store these pruned dense weight blocks to speed up the calculation. In the following experiments on MNIST and Fashion-MNIST datasets, the trend of accuracies with different pruning granularities and different sparsity is explored in order to analyze our method. The experimental results show that our coarse-grained block pruning method can compress the network and can reduce the computational cost without greatly degrading the classification accuracy. The experiment on the CIFAR-10 dataset shows that our block pruning strategy can combine well with the convolutional networks.
20
Cedeno Jimenez, J. R., D. Oxoli, and M. A. Brovelli. "ENABLING AIR QUALITY MONITORING WITH THE OPEN DATA CUBE: IMPLEMENTATION FOR SENTINEL-5P AND GROUND SENSOR OBSERVATIONS." International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLVI-4/W2-2021 (August 19, 2021): 31–36. http://dx.doi.org/10.5194/isprs-archives-xlvi-4-w2-2021-31-2021.
Повний текст джерелаАнотація:
Abstract. Nowadays, the amount of open geospatial data delivered e.g. by private and public entities, such as environmental agencies, enables outstanding possibilities to any user interested in investigating real-world phenomena. However, the availability of such information presents several challenges in terms of its practical use. These are mainly connected to the heterogeneity of data sources, formats and processing tools which have to be mastered by the user to obtain the desired results. As a relevant example, air quality monitoring requires the integration of multiple data with different spatial and temporal granularities that are often distributed by more than one provider using not uniform formats and access methods. Besides traditional air pollution ground sensors observations, novel data sources have emerged. Among them, the Sentinel-5P mission of the European Copernicus Programme is one of the most recent Earth Observation platforms providing estimates of air pollutants with daily global coverage. These estimates are promising to foster air quality analysis and monitoring by complementing ground sensors observations. Therefore, the development of data handling and analysis strategies – allowing users for a smooth integration of satellite and ground sensor observations – is key to support future air quality studies. To that end, the present work investigates the use of the Open Data Cube as a single data endpoint to incorporate ground sensors and satellite observations into local air pollution analyses. A preliminary implementation is presented using the Lombardy region (Northern Italy) as a case study.
21
Gouripeddi, Ram, Andrew Miller, Karen Eilbeck, Katherine Sward, and Julio C. Facelli. "3399 Systematically Integrating Microbiomes and Exposomes for Translational Research." Journal of Clinical and Translational Science 3, s1 (March 2019): 29–30. http://dx.doi.org/10.1017/cts.2019.71.
Повний текст джерелаАнотація:
OBJECTIVES/SPECIFIC AIMS: Characterize microbiome metadata describing specimens collected, genomic pipelines and microbiome results, and incorporate them into a data integration platform for enabling harmonization, integration and assimilation of microbial genomics with exposures as spatiotemporal events. METHODS/STUDY POPULATION: We followed similar methods utilized in previous efforts in charactering and developing metadata models for describing microbiome metadata. Due to the heterogeneity in microbiome and exposome data, we aligned them along a conceptual representation of different data used in translational research; microbiomes being biospecimen-derived, and exposomes being a combination of sensor measurements, surveys and computationally modelled data. We performed a review of literature describing microbiome data, metadata, and semantics [4–15], along with existing datasets [16] and developed an initial metadata model. We reviewed the model with microbiome domain experts for its accuracy and completeness, and with translational researchers for its utility in different studies, and iteratively refined it. We then incorporated the logical model into OpenFurther’s metadata repository MDR [17,18] for harmonization of different microbiome datasets, as well as integration and assimilation of microbiome-exposome events utilizing the UPIE. RESULTS/ANTICIPATED RESULTS: Our model for describing the microbiome currently includes three domains (1) the specimen collected for analysis, (2) the microbial genomics pipelines, and (3) details of the microbiome genomics. For (1), we utilized biospecimen data model that harmonizes the data structures of caTissue, OpenSpecimen and other commonly available specimen management platform. (3) includes details about the organisms, isolate, host specifics, sequencing methodology, genomic sequences and annotations, microbiome phenotype, genomic data and storage, genomic copies and associated times stamps. We then incorporated this logical model into the MDR as assets and associations that UPIE utilizes to harmonize different microbiome datasets, followed by integration and assimilation of microbiome-exposome events. Details of (2) are ongoing. DISCUSSION/SIGNIFICANCE OF IMPACT: The role of the microbiome and co-influences from environmental exposures in etio-pathology of various pulmonary conditions isn’t well understood [19–24]. This metadata model for the microbiome provides a systematic approach for integrating microbial genomics with sensor-based environmental and physiological data, and clinical data that are present in varying spatial and temporal granularities and require complex methods for integration, assimilation and analysis. Incorporation of this microbiome model will advance the performance of sensor-based exposure studies of the (UPIE) to support novel research paradigms that will improve our understanding of the role of microbiome in promoting and preventing airway inflammation by performing a range of hypothesis-driven microbiome-exposome pediatric asthma studies across the translational spectrum.
22
Cheng, Zhifeng, Jianghao Wang, Kaixin Zhu, Yong Ge, and Chenghu Zhou. "Evaluating spatial statistical and machine learning models in urban dynamic population mapping." Transactions in Urban Data, Science, and Technology, August 5, 2022, 275412312211141. http://dx.doi.org/10.1177/27541231221114169.
Повний текст джерелаАнотація:
Understanding population dynamics at fine spatiotemporal granularities are valuable to human-centered studies. With the increasing availability of high-frequency human digital footprint data, the past decades have witnessed numerous efforts in mapping populations at fine spatiotemporal scales. However, such research still lacks a unified standard in modeling strategy and auxiliary data selection, especially a systematic comparison between newly developed machine learning techniques and traditional spatial statistical methods under different covariates provisions. Here, we compared two spatial statistical models, the Bayesian space-time model and geographically and temporally weighted regression, with two machine learning techniques, random forest and eXtreme gradient boosting, in a case study of hourly population mapping at 100 m resolution in Beijing. We evaluated the model performance with varied covariates combinations and found that the Bayesian space-time model achieved the best in conjunction with urban function data. Leveraging the optimal model constructed, we mapped dynamic population distribution and concluded human activity patterns on diverse city amenities. This paper emphasizes the importance of spatiotemporal dependency information in fine temporal scale population mapping and the urban function covariates in urban population mapping.
23
Guillaume, Gwenaël, Arnaud Can, Gwendall Petit, Nicolas Fortin, Sylvain Palominos, Benoit Gauvreau, Erwan Bocher, and Judicaël Picaut. "Noise mapping based on participative measurements." Noise Mapping 3, no. 1 (January 25, 2016). http://dx.doi.org/10.1515/noise-2016-0011.
Повний текст джерелаАнотація:
AbstractThe high temporal and spatial granularities recommended by the European regulation for the purpose of environmental noise mapping leads to consider new alternatives to simulations for reaching such information. While more and more European cities deploy urban environmental observatories, the ceaseless rising number of citizens equipped with both a geographical positioning system and environmental sensors through their smartphones legitimates the design of outsourced systems that promote citizen participatory sensing. In this context, the OnoM@p system aims at offering a framework for capitalizing on crowd noise data recorded by inexperienced individuals by means of an especially designed mobile phone application. The system fully rests upon open source tools and interoperability standards defined by the Open Geospatial Consortium. Moreover, the implementation of the Spatial Data Infrastructure principle enables to break up as services the various business modules for acquiring, analysing and mapping sound levels. The proposed architecture rests on outsourced processes able to filter outlier sensors and untrustworthy data, to cross- reference geolocalised noise measurements with both geographical and statistical data in order to provide higherlevel indicators, and to map the collected and processed data based on web services.
24
Lal, Sohan, Bogaraju Sharatchandra Varma, and Ben Juurlink. "A Quantitative Study of Locality in GPU Caches for Memory-Divergent Workloads." International Journal of Parallel Programming, April 5, 2022. http://dx.doi.org/10.1007/s10766-022-00729-2.
Повний текст джерелаАнотація:
AbstractGPUs are capable of delivering peak performance in TFLOPs, however, peak performance is often difficult to achieve due to several performance bottlenecks. Memory divergence is one such performance bottleneck that makes it harder to exploit locality, cause cache thrashing, and high miss rate, therefore, impeding GPU performance. As data locality is crucial for performance, there have been several efforts to exploit data locality in GPUs. However, there is a lack of quantitative analysis of data locality, which could pave the way for optimizations. In this paper, we quantitatively study the data locality and its limits in GPUs at different granularities. We show that, in contrast to previous studies, there is a significantly higher inter-warp locality at the L1 data cache for memory-divergent workloads. We further show that about 50% of the cache capacity and other scarce resources such as NoC bandwidth are wasted due to data over-fetch caused by memory divergence. While the low spatial utilization of cache lines justifies the sectored-cache design to only fetch those sectors of a cache line that are needed during a request, our limit study reveals the lost spatial locality for which additional memory requests are needed to fetch the other sectors of the same cache line. The lost spatial locality presents opportunities for further optimizing the cache design.
25
Ding, Xuewei, Yingwei Pan, Yehao Li, Ting Yao, Dan Zeng, and Tao Mei. "Boosting Relationship Detection in Images with Multi-Granular Self-Supervised Learning." ACM Transactions on Multimedia Computing, Communications, and Applications, August 18, 2022. http://dx.doi.org/10.1145/3556978.
Повний текст джерелаАнотація:
Visual and spatial relationship detection in images has been a fast-developing research topic in multimedia field, which learns to recognize the semantic/spatial interactions between objects in an image, aiming to compose a structured semantic understanding of the scene. Most of existing techniques directly encapsulate the holistic image feature plus the semantic and spatial features of the given two objects for predicting the relationship, but leave the inherent supervision derived from such structured and thorough image understanding under-exploited. Specifically, the inherent supervision among objects or relations within an image can span different granularities in this hierarchy including, from simple to comprehensive, 1) the object-based supervision that captures the interaction between the semantic and spatial features of each individual object, 2) the inter-object supervision that characterizes the dependency within the relationship triplet ( <subject-predicate-object> ), and 3) the inter-relation supervision that exploits contextual information among all relationship triplets in an image. These inherent multi-granular supervision offers a fertile ground for building self-supervised proxy tasks. In this paper, we compose a trilogy of exploring the multi-granular supervision in the sequence from object-based, inter-object and inter-relation perspectives. We integrate the standard relationship detection objective with a series of proposed self-supervised proxy tasks, which is named as Multi-Granular Self-Supervised learning (MGS). Our MGS is appealing in view that it is pluggable to any neural relationship detection models by simply including the proxy tasks during training, without increasing the computational cost at inference. Through extensive experiments conducted on SpatialSense and VRD datasets, we demonstrate the superiority of MGS for both spatial and visual relationship detection tasks.
26
Jiang, Jing, Ruisheng Zhang, Zhili Zhao, Jun Ma, Yunwu Liu, Yongna Yuan, and Bojuan Niu. "MultiGran-SMILES: Multi-Granularity SMILES Learning for Molecular Property Prediction." Bioinformatics, August 12, 2022. http://dx.doi.org/10.1093/bioinformatics/btac550.
Повний текст джерелаАнотація:
Abstract Motivation Extracting useful molecular features is essential for molecular property prediction. Atom-level representation is a common representation of molecules, ignoring the sub-structure or branch information of molecules to some extent, however, it is vice versa for the substring-level representation. Both atom-level and substring-level representations may lose the neighborhood or spatial information of molecules. While molecular graph representation aggregating the neighborhood information of a molecule has a weak ability in expressing the chiral molecules or symmetrical structure. In this paper, we aim to make use of the advantages of representations in different granularities simultaneously for molecular property prediction. To this end, we propose a fusion model named MultiGran-SMILES, which integrates the molecular features of atoms, sub-structures, and graphs from the input. Compared with the single granularity representation of molecules, our method leverages the advantages of various granularity representations simultaneously and adjusts the contribution of each type of representation adaptively for molecular property prediction. Results The experimental results show that our MultiGran-SMILES method achieves state-of-the-art performance on BBBP, LogP, HIV, and ClinTox datasets. For the BACE, FDA, and Tox21 datasets, the results are comparable with the state-of-the-art models. Moreover, the experimental results show that the gains of our proposed method are bigger for the molecules with obvious functional groups or branches. Availability The code and data underlying this work are available on GitHub, at https://github. com/Jiangjing0122/MultiGran. Supplementary information Supplementary data are available at Bioinformatics online.
27
Li, Mingke, and Emmanuel Stefanakis. "Geo-feature modeling uncertainties in discrete global grids: a case study of downtown Calgary, Canada." Geomatica, October 26, 2020, 1–21. http://dx.doi.org/10.1139/geomat-2020-0011.
Повний текст джерелаАнотація:
The Open Geospatial Consortium has officially adopted discrete global grid systems (DGGS) as a new option for Earth reference standards. Many state-of-the-art DGGS implementations have been developed, revealing the potential for DGGS applications. Before the wide application of DGGS in solving real-world problems, however, the potential uncertainties of modeling on DGGS should be investigated and documented. This study focused on the uncertainties of geo-feature modeling on DGGS, quantitatively measured the point position displacement and line and polygon features’ geometry distortion, and evaluated the validity of topological relationships. Specifically, traffic cameras (points), main streets (lines), and land-cover classes (polygons) of downtown Calgary (AB, Canada) were modeled in various DGGS configurations at multiple resolutions. Results showed that the point displacement and polygon distortion generally reduced when being modeled at a higher resolution. The tessellations with the monotonical convergence characteristic are recommended if cell indices are expected to represent levels of model precision. Line features’ fidelity was affected by grid tessellations, resolution levels, grid orientation relative to the Earth, and the rotated line directions. The degree of the line distortion was not straightforward to forecast. Maintaining the topological validity between spatial objects with various granularities was challenging and needed further algorithm development for DGGS implementations. The study outcomes can serve as useful guidelines in the selection among grid types, refinement ratios, and resolution levels when applying DGGS implementations to urban environments. This paper also pinpoints several research directions that can benefit the quantization and analysis of vector features on DGGS.