No abstract

Car-following model is indispensable to evaluate the characteristics of car-following behaviors. Through an analysis and comparison of data-driven and theoretically driven car-following models, it shows that the data-driven model has poor interpretability and high quality of training data set is required, while for the theoretical-driven model, it is unable to describe the individualized features and models of the driver so as to a low model accuracy. To solve the problem, a novel modelling method is proposed using adaptive Kalman filter algorithm to integrate the long-short-time memory neural network (LSTM) data-driven model with the IDM theoretical-driven model to build the car-following model. Test results of real driving data from a single driver prove that the fusion car-following model has higher accuracy than a single model, while at the same time highlighting the driver's personality compared to the IDM model. Besides, it improves the generalization ability of the traditional data model, which is reflected by better fitting in the extreme case (for example, the stable state when the acceleration, velocity is zero). Finally, the trajectory simulation results show that the proposed integrated data-driven car-following model can better simulate the micro-traffic behavior of car following.INDEX TERMS Car-following characteristic, LSTM, IDM, adaptive Kalman filter, fusion modelling.

Face age estimation has been widely used in video surveillance, human-computer interaction, market analysis, image processing analysis, and many fields. There are several problems that need to be solved in image-based face age estimation: (1) redundant information of age characteristics; (2) limitations of age estimation methods in solving age estimation problems; (3) the performance of age estimation models being also affected by gender factors. This paper proposes CA-XTree network. Firstly, features are extracted through the convolution layer and then combined with the local channel attention module to strengthen the ability of age feature information interaction between different channels. Secondly, extracted features are inputted into the recommendation score function to obtain the recommendation score, by combining the recommendation score with the gradient ascending regression tree. The lifting tree processed loss function is the mean square loss function, and the final age value is obtained by the leaf node. This paper improves state of the art for image classification on MORPH and CACD datasets. The advantage of our model is that it is easy to implement and has no excess memory overhead. In the age dataset CACD, the mean absolute error (MAE) has reached 4.535 and cumulative score (CS) has reached 63.53%, respectively.