Origin-destination (OD) flow pattern mining is an important research method of urban dynamics, in which OD flow clustering analysis discovers the activity patterns of urban residents and mine the coupling relationship of urban subspace and dynamic causes. The existing flow clustering methods are limited by the spatial constraints of OD points, rely on the spatial similarity of geographical points, and lack in-depth analysis of high-dimensional flow characteristics, and therefore it is difficult to find irregular flow clusters. In this paper, we propose an OD flow clustering method based on vector constraints (ODFCVC), which defines OD flow event point and OD flow vector to express the spatial location relationship and geometric flow behavior characteristics of OD flow. First, the OD flow vector coordinate system is normalized by the Euclidean distance-based OD flow event point spatial clustering, and then the OD flow clusters with similar flow patterns are mined using adjusted cosine similarity-based OD flow vector feature clustering. The transformation of OD data from point set space to vector space is realized by constraining the vector coordinate system and vector similarity through two-step clustering, which simplifies the calculation of high-dimensional similarity of OD flow and helps mining representative OD flow clusters in flow space. Due to the OD flow cluster property, the k-means algorithm is selected as the basic clustering logic in the two-step clustering method, and a sum of squared error perceptually important points algorithm considering silhouette coefficients (SSEPIP) is adopted to automatically extract the optimal cluster number without defining any parameters. Tested by origin-destination flow data in Beijing, China, new traffic flow communities based on traffic hubs are obtained by using the ODFCVC method, and irregular traffic flow clusters (including cluster mode, divergence mode, and convergence mode) with representative travel trends are found.

The prediction and control of passenger flow in scenic spots is very important to the traffic management and safety of scenic spots. This study aims to predict the passenger flow of a scenic spot based on the passenger flow of the bus and subway stations around the scenic spots. We propose a passenger flow prediction model based on graph convolutional network–recurrent neural network (GCN–RNN). First, a “graph” is constructed according to the geographical relationship between the scenic spot and the surrounding bus and subway stations. Then, characteristics of surrounding areas of bus and subway stations are constructed based on the crowd behavior analysis, and these are then used as the node-information of the “graph”. Last, the GCN–RNN model is used to extract the temporal and spatial characteristics of the passenger flow data of the scenic spot to realize the prediction. The experimental results show that the proposed model is effective in passenger flow prediction in scenic spots.

Lane detection is a crucial task in the field of autonomous driving, as it enables vehicles to safely navigate on the road by interpreting the high-level semantics of traffic signs. Unfortunately, lane detection is a challenging problem due to factors such as low-light conditions, occlusions, and lane line blurring. These factors increase the perplexity and indeterminacy of the lane features, making them hard to distinguish and segment. To tackle these challenges, we propose a method called low-light enhancement fast lane detection (LLFLD) that integrates the automatic low-light scene enhancement network (ALLE) with the lane detection network to improve lane detection performance under low-light conditions. Specifically, we first utilize the ALLE network to enhance the input image’s brightness and contrast while reducing excessive noise and color distortion. Then, we introduce symmetric feature flipping module (SFFM) and channel fusion self-attention mechanism (CFSAT) to the model, which refine the low-level features and utilize more abundant global contextual information, respectively. Moreover, we devise a novel structural loss function that leverages the inherent prior geometric constraints of lanes to optimize the detection results. We evaluate our method on the CULane dataset, a public benchmark for lane detection in various lighting conditions. Our experiments show that our approach surpasses other state of the arts in both daytime and nighttime settings, especially in low-light scenarios.

With the rapid growth of trajectory big data, there is a need for more efficient methods to extract, analyze, and visualize these data. However, existing research on trajectory big data visualization mainly focuses on displaying trajectories for a specific period or showing spatial distribution characteristics of trajectory points in a single time slice using clustering, filtering, and other techniques. Therefore, this paper proposes a vector field visualization model for trajectory big data, aiming to effectively represent the inherent movement trends in the data and provide a more intuitive visualization of urban traffic congestion trends. The model utilizes the motion information of vehicles to create a travel vector grid and employs WebGL technology for vector field visualization rendering. The vector field effects are effectively displayed by generating many particles and simulating their movements. Furthermore, this research also designs and implements congestion trend point identification and hotspot congestion analysis, thus validating the practicality and effectiveness of trajectory big data vector field visualization. The results indicate that compared to traditional visualization methods, the vector field visualization method can demonstrate the direction and density changes in traffic flow and predict future traffic congestion. This work provides valuable data references and decision support for urban traffic management and planning.

The fashion industry has a critical need for fashion compatibility. Modeling compatibility is a challenging task that involves extracting (in)compatible features of pairs, obtaining compatible relationships between matching items, and applying them to personalized recommendation tasks. Measuring compatibility is a complex and subjective concept in general. The complexity is reflected in the fact that relationships between fashion items are determined by multiple matching rules, such as color, shape, and material. Each personal aesthetic style and fashion preference differs, adding subjectivity to the compatibility concept. As a result, personalized factors must be considered. Previous works mainly utilize a convolutional neural network to measure compatibility by extracting general features, but they ignore fine-grained compatibility features and only model overall compatibility. We propose a novel neural network framework called the Attention-based Personalized Compatibility Embedding Network (PCE-Net). It comprises two components: attention-based compatibility embedding modeling and attention-based personal preference modeling. In the second part, we utilize matrix factorization and content-based features to obtain user preferences. Both pieces are jointly trained using the BPR framework in an end-to-end method. Extensive experiments on the IQON3000 dataset demonstrate that PCE-Net significantly outperforms most baseline methods.