The convolutional neural network (CNN) has been widely used in image recognition field due to its good performance. This paper proposes a facial expression recognition method based on the CNN model. Regarding the complexity of the hierarchic structure of the CNN model, the activation function is its core, because the nonlinear ability of the activation function really makes the deep neural network have authentic artificial intelligence. Among common activation functions, the ReLu function is one of the best of them, but it also has some shortcomings. Since the derivative of the ReLu function is always zero when the input value is negative, it is likely to appear as the phenomenon of neuronal necrosis. In order to solve the above problem, the influence of the activation function in the CNN model is studied in this paper. According to the design principle of the activation function in CNN model, a new piecewise activation function is proposed. Five common activation functions (i.e., sigmoid, tanh, ReLu, leaky ReLus and softplus–ReLu, plus the new activation function) have been analysed and compared in facial expression recognition tasks based on the Keras framework. The Experimental results on two public facial expression databases (i.e., JAFFE and FER2013) show that the convolutional neural network based on the improved activation function has a better performance than most-of-the-art activation functions.

The aspect ratio of a target changes frequently during an unmanned aerial vehicle (UAV) tracking task, which makes the aerial tracking very challenging. Traditional trackers struggle from such a problem as they mainly focus on the scale variation issue by maintaining a certain aspect ratio. In this paper, we propose a coarse-to-fine deep scheme to address the aspect ratio variation in UAV tracking. The coarse-tracker first produces an initial estimate for the target object, then a sequence of actions are learned to fine-tune the four boundaries of the bounding box. The coarse-tracker and the fine-tracker are designed to have different action spaces and operating target. The former dominates the entire bounding box and the latter focuses on the refinement of each boundary. They are trained jointly by sharing the perception network with an end-to-end reinforcement learning architecture. Experimental results on benchmark aerial data set prove that the proposed approach outperforms existing trackers and produces significant accuracy gains in dealing with the aspect ratio variation in UAV tracking.

The performance of vehicle body vibration and ride comfort of active or semi-active suspension with proper control is better than that with passive suspension. The key to achieve good control effect is that the suspension control system should have strong real-time learning ability according to changes in the road surface and suspension parameters. In the control strategies adopted by previous researchers, the classical neural network controller has some learning ability, but it is mainly based on offline learning with a large number of samples. In this paper, the deep reinforcement learning strategy is used to solve the above problems. Aiming at the continuity of state space and execution action in vehicle active suspension system, the control of the semi-active suspension is realized by using improved DDPG (Deep Deterministic Policy Gradient) algorithm. To overcome the shortcoming of low efficiency of this algorithm in the initial stage of learning, the DDPG algorithm is improved and using empirical samples in the learning method is proposed. Based on Mujoco, the physical model of semi-active suspension is established, and its dynamic characteristics are analyzed under the condition of various road level and vehicle speed. The simulation results show that compared with the passive suspension, the semi-active suspension based on improved DDPG algorithm with learning method using experienced samples can better adapt to various road level, more effectively reduce the vertical acceleration of the vehicle body and the dynamic deflection of the suspension, and further improve the ride comfort.

In this study, an autonomous robot navigation system is designed for live working on distribution line. The developed system features a real‐time detection and motion planning system, incorporating a manipulator capable of grasping power components. In order to accurately identify targets, the authors propose an object detection method based on the Larger Scale ‘You Only Look Once’ Version 4 (LS‐YOLOv4) algorithm for detecting the insulators and drop fuses. The LS‐YOLOv4 extracts features of power components by Convolutional Neural Network (CNN), and then performs feature fusion. Then the authors develop a motion planning method based on the Node Control Optimal Rapidly Exploring Random Trees (NC‐RRT*), which can drive the robot to realise the autonomous robot motion planning and obstacle avoidance. On the grasping function, the authors present a reliable Lightweight‐based Convolutional Neural Network (L‐CNN) grasping point detection method. Finally, the authors evaluate fully autonomous robotic system in both simulated and real‐world experiments. The experimental results demonstrate that the proposed system can effectively identify the target and complete the grasping task in an efficient way. Notably, the proposed motion planning method can take into account both planning efficiency and accuracy to manipulation tasks.

As an important part of emotion research, facial expression recognition is a necessary requirement in human–machine interface. Generally, a face expression recognition system includes face detection, feature extraction, and feature classification. Although great success has been made by the traditional machine learning methods, most of them have complex computational problems and lack the ability to extract comprehensive and abstract features. Deep learning-based methods can realize a higher recognition rate for facial expressions, but a large number of training samples and tuning parameters are needed, and the hardware requirement is very high. For the above problems, this paper proposes a method combining features that extracted by the convolutional neural network (CNN) with the C4.5 classifier to recognize facial expressions, which not only can address the incompleteness of handcrafted features but also can avoid the high hardware configuration in the deep learning model. Considering some problems of overfitting and weak generalization ability of the single classifier, random forest is applied in this paper. Meanwhile, this paper makes some improvements for C4.5 classifier and the traditional random forest in the process of experiments. A large number of experiments have proved the effectiveness and feasibility of the proposed method.