Spatiotemporal Co-occurrence Patterns Research Articles

Progressive unsupervised video person re-identification with accumulative motion and tracklet spatial–temporal correlation

Person re-identification (re-id) methods based on supervised learning require a large number of manually labeled samples for training, resulting in poor scalability in actual re-id applications. Existing unsupervised video person re-id methods typically focus on extracting appearance features from pedestrian videos, ignoring motion information and the fact that people usually move in groups, i.e., pedestrian spatio–temporal co-occurrence patterns. The key factor for unsupervised video person re-id is to effectively exploit both spatio–temporal clues from video sequences and cross-camera tracklet association. In this work, we propose a progressive deep learning method for unsupervised person re-id via tracklet association with spatio–temporal correlation (TASTC). Specifically, we first divide uniformly each tracklet into multiple temporally localized slices according to a time pyramid structure. Then, an initial re-id model is trained based on a two-stream convolutional architecture, and the accumulative motion context information of temporally localized slices of the tracklets per camera is learned. Finally, combining accumulative motion and tracklet spatial–temporal correlation, we associate tracklets across cameras and update the re-id model. The above steps are iterated to optimize the re-id model progressively. Experiment results demonstrate that the proposed method significantly outperforms the current state-of-the-art unsupervised video person re-identification methods on three video-based benchmark datasets, ILIDS-VID, MARS, and DukeMTMC-VideoReID.

SPATIOTEMPORAL CO-OCCURRENCE PATTERNS OF DEMERSAL FISHES IN A SUBTROPICAL MARINE AND ESTUARINE ENVIRONMENT

In order to investigate spatiotemporal co-occurrence patterns of demersal fishes from the Paranaguá Estuarine Complex (PEC) and the adjacent continental shelf, state of Paraná, Southern Brazil, fish from nine sampling sites within the estuary and three additional sites in open sea (N = 12 sampling sites) were sampled at quarterly intervals between 2014 and 2016. The assemblage structure was evaluated by the C-Score index and null models developed through 5,000 randomizations. Species co-occurred as expected, but a general pattern in both temporal and spatial matrices was only detected in the Upper sector, which showed a random co-occurrence pattern. In contrast, fish species from the other sectors co-occurred less than expected by chance in some circumstances. According to the concepts related to the organization of communities on temporal scales, many communities tend to become disorganized after disturbances and then present a progressive increase in organization, that is, they display a random co-occurrence pattern, similar to initial response, and over time the species in the community segregate or aggregate in response to system stability restoration. Therefore, understanding how fish assemblages are structured at different spatial and temporal scales, and especially what processes drive such conformations, including assembly rules for communities, is crucial for ecosystem maintenance and understanding, as well as providing baselines for conservation and management policies, interventions, and practices in these environments.

Heterogeneous spatio-temporal relation learning network for facial action unit detection

• The Attention-based Feature Decoupling module is proposed to obtain AU-specific spatio-temporal feature vectors. • The Co-occurrence Knowledge Graph module is proposed to introduce the prior spatio-temporal co-occurrence between AUs. • The Spatio-Temporal Transformer is proposed to obtain the spatio-temporal cooccurrence pattern in a data-driven manner. • An end-to-end facial AU detection network based on heterogeneous spatiotemporal relation learning is proposed. Properties such as temporal relation and action relation of facial Action Units (AUs) make AU detection different from general multi-label classification tasks. Therefore, how to capture the spatial and temporal co-occurrence of AUs becomes the key to improving detection accuracy. Although many works on the spatial relation of AUs have been proposed in recent years, very few works have explored the temporal relation of AUs. In this paper, we propose a Heterogeneous Spatio-Temporal Relation learning network (HSTR-Net) to capture the temporal and spatial relations of AUs. The co-occurrence knowledge graph module guides the network to model spatio-temporal relations by introducing prior relation information, and the spatio-temporal Transformer module adaptively captures spatio-temporal relations through spatio-temporal feature interaction. To further model the temporal relation between AUs, the self-attention mechanism to fuse AU features in the time dimension. Extensive experiments are conducted on two challenging datasets, BP4D and DISFA, and experimental results show that our proposed HSTR-Net achieves comparable performance of the state-of-the-art in the field of AU detection. The code for our methods is available at https://github.com/saaaaalmon/HSTR-Net .

Discovering Spatio-Temporal Co-Occurrence Patterns of Crimes with Uncertain Occurrence Time

The discovery of spatio-temporal co-occurrence patterns (STCPs) among multiple types of crimes whose events frequently co-occur in neighboring space and time is crucial to the joint prevention of crimes. However, the crime event occurrence time is often uncertain due to a lack of witnesses. This occurrence time uncertainty further results in the uncertainty of the spatio-temporal neighborhood relationships and STCPs. Existing methods have mostly modeled the uncertainty of events under the independent and identically distributed assumption and utilized one-sided distance information to measure the distance between uncertain events. As a result, STCPs detected from a dataset with occurrence time uncertainty (USTCPs) are likely to be erroneously assessed. Therefore, this paper proposes a probabilistic-distance-based USTCP discovery method. First, the temporal probability density functions of crime events with uncertain occurrence times are estimated by considering the temporal dependence. Second, the spatio-temporal neighborhood relationships are constructed based on the spatial Euclidean distance and the proposed temporal probabilistic distance. Finally, the prevalent USTCPs are identified. Experimental comparisons performed on twelve types of crimes from X City Public Security Bureau in China demonstrate that the proposed method can more objectively express the occurrence time of crimes and more reliably identify USTCPs.

A Spatio-Temporal Co-Clustering Framework for Discovering Mobility Patterns: A Study of Manhattan Taxi Data

Research on clustering spatio-temporal data to extract mobility patterns requires further development, as most existing studies do not simultaneously integrate data along both spatial dimensions and temporal dimensions but instead focus on only one dimension or separate the dimensions in analyses and applications, which could lead to discoveries that are not representative of the overall data or are dificult to interpret. To simultaneously reveal the spatial and temporal patterns of urban mobility datasets, we propose an analytical framework that is based on co-clustering and enables mobility behaviors to be distinguished in spatial and temporal dimensions. We use one month of taxi GPS data from the Manhattan area to explore spatio-temporal co-occurrence patterns. The spatial and temporal dimensions of taxi trip data were co-clustered by using the Bregman Block Average co-clustering algorithm with I-divergence (BBAC_I). We performed this process on weekdays and holidays and compared the mobility differences between these two periods. The experimental results demonstrated the effectiveness of this analytical framework, with which we can reveal the spatial patterns and their temporal dynamics as well as temporal patterns and their spatial dynamics in mobility data.

Towards Better Detection and Analysis of Massive Spatiotemporal Co-Occurrence Patterns

With the rapid development of sensing technologies, massive spatiotemporal data have been acquired from the urban space with respect to different domains, such as transportation and environment. Numerous co-occurrence patterns (e.g., traffic speed <; 10km/h, weather = foggy, and air quality = unhealthy) between the transportation data and other types of data can be obtained with given spatiotemporal constraints (e.g., within 3 kilometers and lasting for 2 hours) from these heterogeneous data sources. Such patterns present valuable implications for many urban applications, such as traffic management, pollution diagnosis, and transportation planning. However, extracting and understanding these patterns is beyond manual capability because of the scale, diversity, and heterogeneity of the data. To address this issue, a novel visual analytics system called CorVizor is proposed to identify and interpret these co-occurrence patterns. CorVizor comprises two major components. The first component is a co-occurrence mining framework involving three steps, namely, spatiotemporal indexing, co-occurring instance generation, and pattern mining. The second component is a visualization technique called CorView that implements a level-of-detail mechanism by integrating tailored visualizations to depict the extracted spatiotemporal co-occurrence patterns. The case studies and expert interviews are conducted to demonstrate the effectiveness of CorVizor.

A statistical method for detecting spatiotemporal co-occurrence patterns

ABSTRACTSpatiotemporal co-occurrence patterns (STCOPs) are subsets of Boolean features whose instances frequently co-occur in both space and time. The detection of STCOPs is crucial to the investigation of the spatiotemporal interactions among different features. However, prevalent STCOPs reported by available methods do not necessarily indicate the statistically significant dependence among different features, which is likely to result in highly erroneous assessments in practice. To improve the reliability of results, this paper develops a statistical method to detect STCOPs and discern their statistical significance. The proposed method detects STCOPs against the null hypothesis that the spatiotemporal distributions of different features are independent of each other. To construct the null hypothesis, suitable spatiotemporal point-process models considering spatiotemporal autocorrelation are employed to model the distributions of different features. The performance of the proposed statistical method is assessed by synthetic experiments and a case study aimed at identifying crime patterns among multiple crime types in Portland City. The experimental results demonstrate that the proposed method is more effective for detecting meaningful STCOPs than the available alternative methods.

Research of MDCOP mining based on time aggregated graph for large spatio-temproal data sets

Discovering mixed-drove spatiotemporal co-occurrence patterns (MDCOPs) is important for network security such as distributed denial of service (DDoS) attack. There are usually many features when we are suffering from a DDoS attacks such as the server CPU is heavily occupied for a long time, bandwidth is hoovered and so on. In distributed cooperative intrusion, the feature information from multiple intrusion detection sources should be analyzed simultaneously to find the spatial correlation among the feature information. In addition to spatial correlation, intrusion also has temporal correlation. Some invasions are gradually penetrating, and attacks are the result of cumulative effects over a period of time. So it is necessary to discover mixed-drove spatiotemporal co-occurrence patterns (MDCOPs) in network security. However, it is difficult to mine MDCOPs from large attack event data sets because mining MDCOPs is computationally very expensive. In information security, the set of candidate co-occurrence attack event data sets is exponential in the number of object-types and the spatiotemporal data sets are too large to be managed in memory. To reduce the number of candidate co-occurrence instances, we present a computationally efficient MDCOP Graph Miner algorithm by using Time Aggregated Graph. which can deal with large attack event data sets by means of file index. The correctness, completeness and efficiency of the proposed methods are analyzed.

Vessel Spatio-temporal Knowledge Discovery with AIS Trajectories Using Co-clustering

Large volumes of data collected by the Automatic Identification System (AIS) provide opportunities for studying both single vessel motion behaviours and collective mobility patterns on the sea. Understanding these behaviours or patterns is of great importance to maritime situational awareness applications. In this paper, we leveraged AIS trajectories to discover vessel spatio-temporal co-occurrence patterns, which distinguish vessel behaviours simultaneously in terms of space, time and other dimensions (such as ship type, speed, width etc.). To this end, available AIS data were processed to generate spatio-temporal matrices and spatio-temporal tensors (i.e., multidimensional arrays). We then imposed a sparse bilinear decomposition on the matrices and a sparse multi-linear decomposition on the tensors. Experimental results on a real-world dataset demonstrated the effectiveness of this methodology, with which we show the existence of connection among regions, time, and vessel attributes.

Mining At Most Top-K% Spatiotemporal Co-occurrence Patterns in Datasets with Extended Spatial Representations

Spatiotemporal co-occurrence patterns (STCOPs) in datasets with extended spatial representations are two or more different event types, represented as polygons evolving in time, whose instances often occur together in both space and time. Finding STCOPs is an important problem in domains such as weather monitoring, wildlife migration, and solar physics. Nevertheless, in real life, it is difficult to find a suitable prevalence threshold without prior domain-specific knowledge. In this article, we focus our work on the problem of mining at most top-K% of STCOPs from continuously evolving spatiotemporal events that have polygon-like representations, without using a user-specified prevalence threshold.

Mining spatiotemporal co-occurrence patterns in non-relational databases

Spatiotemporal co-occurrence patterns (STCOPs) represent the subsets of feature types whose instances are frequently co-occurring both in space and time. Spatiotemporal co-occurrences reflect the spatiotemporal overlap relationships among two or more spatiotemporal instances both in spatial and temporal dimensions. STCOPs can be potentially used to predict and understand the generation and evolution of different types of interacting phenomena in various scientific fields such as astronomy, meteorology, biology, geosciences. Meaningful and statistically significant data analysis for these scientific fields requires processing sufficiently large datasets. Due to the computationally expensive nature of spatiotemporal operations required for mining spatiotemporal co-occurrences, it is increasingly difficult to identify spatiotemporal co-occurrences and discover STCOPs in centralized system settings. As a solution, we developed a cloud-based distributed mining system for discovering STCOPs. Our system uses Accumulo, a column-oriented non-relational database management system as its backbone. In order to efficiently mine the STCOPs, we propose three data models for managing trajectory-based spatiotemporal data in Accumulo. We introduce an in-memory join-index structure and a join algorithm for effectively performing spatiotemporal join operations on spatiotemporal trajectories in non-relational databases. Lastly, with the experiments with artificial and real life datasets, we evaluate the performance of the proposed models for STCOP mining.

Mining spatiotemporal co-occurrence patterns in solar datasets

Abstract We address the problem of mining spatiotemporal co-occurrence patterns (STCOPs) in solar datasets with extended polygon-based geometric representations. Specifically designed spatiotemporal indexing techniques are used in the mining of STCOPs. These include versions of two well-known spatiotemporal trajectory indexing techniques: the scalable and efficient trajectory index and Chebyshev polynomial indexing. We present a framework, Stcop-Miner , implementing a filter-and-refine STCOP mining algorithm, with the indexing techniques mentioned for efficiently performing data analysis.

Identification of air pollution patterns using a modified fuzzy co-occurrence pattern mining method

Spatio-temporal co-occurrence patterns represent subsets of object types which are located together in both space and time. Discovering spatio-temporal co-occurrence patterns is an important task having many application domains. There are a number of developed methods to mine co-occurrence patterns; however, using them needs a unique parameter to define the neighborhood. Identification of a unique optimum k-value or neighborhood radius is a challenging issue in different application domains. The developed method of this research defines a new fuzzy neighborhood and new fuzzy metrics to be applicable for real applications such as air pollution, especially when the researchers have no comprehensive knowledge regarding the application domain; in addition, it mines patterns based on the fuzzy nature of environmental phenomena. The new method mines patterns locally without localization step to speed up the mining process and considers all feature types (point, line and polygon) to handle all applications. Subsequently, it is applied to a real data set of Tehran city for air pollution to discover significant co-occurrence patterns of air pollution and influencing environmental parameters such as meteorological, topography and traffic. The case study results showed seven meaningful patterns among air pollution classes 2 and 3 and wind speed class 1, topography class 1 and traffic classes 1 and 2. The evaluation confirmed the accuracy and applicability of the new developed method for air pollution case. Furthermore, two performance tests for the method itself and a performance test against a crisp method were done, where the results exhibited an efficient computational performance.

A generic regional spatio-temporal co-occurrence pattern mining model: a case study for air pollution

Spatio-temporal co-occurrence patterns represent subsets of object types which are located together in both space and time. Existing algorithms for co-occurrence pattern mining cannot handle complex applications such as air pollution in several ways. First, the existing models assume that spatial relationships between objects are explicitly represented in the input data, while the new method allows extracting implicitly contained spatial relationships algorithmically. Second, instead of extracting co-occurrence patterns of only point data, the proposed method deals with different feature types that is with point, line and polygon data. Thus, it becomes relevant for a wider range of real applications. Third, it also allows mining a spatio-temporal co-occurrence pattern simultaneously in space and time so that it illustrates the evolution of patterns over space and time. Furthermore, the proposed algorithm uses a Voronoi tessellation to improve efficiency. To evaluate the proposed method, it was applied on a real case study for air pollution where the objective is to find correspondences of air pollution with other parameters which affect this phenomenon. The results of evaluation confirm not only the capability of this method for co-occurrence pattern mining of complex applications, but also it exhibits an efficient computational performance.

Partial spatio-temporal co-occurrence pattern mining

Spatio-temporal co-occurrence patterns represent subsets of object-types that are often located together in space and time. The aim of the discovery of partial spatio-temporal co-occurrence patterns (PACOPs) is to find co-occurrences of the object-types that are partially present in the database. Discovering PACOPs is an important problem with many applications such as discovering interactions between animals in ecology, identifying tactics in battlefields and games, and identifying crime patterns in criminal databases. However, mining PACOPs is computationally very expensive because the interest measures are computationally complex, databases are larger due to the archival history, and the set of candidate patterns is exponential in the number of object-types. Previous studies on discovering spatio-temporal co-occurrence patterns do not take into account the presence period (i.e., lifetime) of the objects in the database. This paper defines the problem of mining PACOPs, proposes a new monotonic composite interest measure, and proposes novel PACOP mining algorithms. The experimental results show that the proposed algorithms are computationally more efficient than the naive alternatives.

Mixed-Drove Spatiotemporal Co-Occurrence Pattern Mining

Mixed-drove spatio-temporal co-occurrence patterns (MDCOPs) represent subsets of two or more different object-types whose instances are often located in spatial and temporal proximity. Discovering MDCOPs is an important problem with many applications such as identifying tactics in battlefields, games, and predator-prey interactions. However, mining MDCOPs is computationally very expensive because the interest measures are computationally complex, datasets are larger due to the archival history, and the set of candidate patterns is exponential in the number of object-types. We propose a monotonic composite interest measure for discovering MDCOPs and novel MDCOP mining algorithms. Analytical results show that the proposed algorithms are correct and complete. Experimental results also show that the proposed methods are computationally more efficient than naive alternatives.

Crime and Punishment: An Economic Approach

Since the turn of the twentieth century, legislation in Western countries has expanded rapidly to reverse the brief dominance of laissez faire during the nineteenth century. The state no longer merely protects against violations of person and property through murder, rape, or burglary but also restricts ‘discrimination’ against certain minorities, collusive business arrangements, ‘jaywalking’, travel, the materials used in construction, and thousands of other activities. The activities restricted not only are numerous but also range widely, affecting persons in very different pursuits and of diverse social backgrounds, education levels, ages, races, etc. Moreover, the likelihood that an offender will be discovered and convicted and the nature and extent of punishments differ greatly from person to person and activity to activity. Yet, in spite of such diversity, some common properties are shared by practically all legislation, and these properties form the subject matter of this essay.