A vehicle logo occupies a small proportion of a car and has different shapes. These characteristics bring difficulties to machine-vision-based vehicle logo detection. To improve the accuracy of vehicle logo detection in complex backgrounds, an improved YOLOv4 model was presented. Firstly, the CSPDenseNet was introduced to improve the backbone feature extraction network, and a shallow output layer was added to replenish the shallow information of small target. Then, the deformable convolution residual block was employed to reconstruct the neck structure to capture the various and irregular shape features. Finally, a new detection head based on a convolutional transformer block was proposed to reduce the influence of complex backgrounds on vehicle logo detection. Experimental results showed that the average accuracy of all categories in the VLD-45 dataset was 62.94%, which was 5.72% higher than the original model. It indicated that the improved model could perform well in vehicle logo detection.
Traffic flow prediction based on spatial–temporal data plays a vital role in traffic management. However, it still faces serious challenges due to the complex spatial–temporal correlation in nonlinear spatial–temporal data. Some previous methods have limited ability to capture spatial–temporal correlation, and ignore the quadratic complexity problem in the traditional attention mechanism. To this end, we propose a novel spatial–temporal combination and multi-head flow-attention network (STCMFA) to model the spatial–temporal correlation in road networks. Firstly, we design a temporal sequence multi-head flow attention (TS-MFA), in which the unique source competition mechanism and sink allocation mechanism make the model avoid attention degradation without being affected by inductive biases. Secondly, we use GRU instead of the linear layer in traditional attention to map the input sequence, which further enhances the temporal modeling ability of the model. Finally, we combine the GCN with the TS-MFA module to capture the spatial–temporal correlation, and introduce residual mechanism and feature aggregation strategy to further improve the performance of STCMFA. Extensive experiments on four real-world traffic datasets show that our model has excellent performance and is always significantly better than other baselines.
Accurately predicting traffic flow is paramount for the efficient operation of transportation systems. The key to enhancing prediction accuracy lies in effectively mining the intricate spatio-temporal correlations within traffic flow data. However, traditional traffic flow prediction methods that combine Graph Convolutional Network and Recurrent Neural Network have limitations in capturing comprehensive spatial correlation information and face challenges in modeling long-term temporal dependencies, consequently leading to suboptimal prediction performance. This study proposes a hybrid traffic flow prediction model based on fusion graph convolutional network and enhanced gate recurrent unit. Initially, a fusion graph structure is constructed based on adjacency graph and adaptive graph to better represent the correlations between nodes in the road network. Subsequently, the stacked fusion graph convolution module is utilized to capture multi-level spatial correlations and the enhanced gated recurrent unit is applied to extract multi-scale temporal correlations. In addition, the model integrates the extracted spatio-temporal features with the direct features through residual connection units, and utilizes the fused features for prediction, achieving superior predictive performance. The experimental results from four authentic datasets demonstrate that our proposed model outperforms state-of-the-art baseline models, showcasing an average enhancement of 3% in Mean Absolute Error(MAE), 3.3% in Root Mean Square Error(RMSE), and 2.7% in Mean Absolute Percentage Error(MAPE) across the four datasets.
Although data-independent acquisition (DIA) has the ability to identify and quantify all peptides in a sample, highly complex mixed mass spectra present difficulties for accurate peptide and protein identification. Additionally, the correspondence between the precursor and its fragments is broken, making it challenging to perform peptide identification directly using conventional DDA search engines. In this paper, we propose a cosine-similarity-based deconvolution method: CorrDIA. This is achieved by reconstructing the correspondence between precursor and fragment ions based on the consistency of extracted ion chromatograms (XICs). A deisotope peak cluster operation is added and centered on the MS/MS spectrum to improve the accuracy of spectrum interpretation and increase the number of identified peptides. The resulting MS/MS spectra can be identified using any data-dependent acquisition (DDA) sequencing software. The experimental results demonstrate that the number of peptide results increased by 12 percent and 21 percent respectively, and the repetition rate decreased by 12 percent. This reduces mass spectra complexity and difficulties in mass spectra analysis without the need for any mass spectra libraries.
4 June 2025 in Engineering Applications of Artificial Intelligence