Search

Fatigued driving is one of the leading causes of traffic accidents, and detecting fatigued driving effectively is critical to improving driving safety. Given the variety and individual variability of the driving surroundings, the drivers’ states of weariness, and the uncertainty of the key characteristic factors, in this paper, we propose a deep-learning-based study of the MAX-MIN driver fatigue detection algorithm. First, the ShuffleNet V2K16 neural network is used for driver face recognition, which eliminates the influence of poor environmental adaptability in fatigue detection; second, ShuffleNet V2K16 is combined with Dlib to obtain the coordinates of driver face feature points; and finally, the values of EAR and MAR are obtained by comparing the first 100 frames of images to EAR-MAX and MAR-MIN. Our proposed method achieves 98.8% precision, 90.2% recall, and 94.3% F-Score in the actual driving scenario application.

Aiming at the problems of frequent identity switches (IDs) and trajectory interruption of multi-pedestrian tracking algorithms in dense scenes, this paper proposes a multi-pedestrian tracking algorithm based on an attention mechanism and dual data association. First, the FairMOT algorithm is used as a baseline to introduce the feature pyramid network in the CenterNet detection network and up-sampling the output multi-scale fused feature maps, effectively reducing the rate of missed detection of small-sized and obscured pedestrians. The improved channel attention mechanism module is embedded in the CenterNet’s backbone network to improve detection accuracy. Then, a re-identification (ReID) branch is embedded in the head of the detection network, and the two sub-tasks of pedestrian detection and pedestrian apparent feature extraction are combined in a multi-task joint learning approach to output the pedestrian apparent feature vectors while detecting pedestrians, which improves the computational efficiency and localization accuracy of the algorithm. Finally, we propose a dual data association tracking model that tracks by associating almost every detection box instead of only the high-scoring ones. For low-scoring detection boxes, we utilize their similarities with trajectories to recover obscured pedestrians. The experiment using the MOT17 dataset shows that the tracking accuracy is improved by 0.6% compared with the baseline FairMOT algorithm, and the number of switches decreases from 3303 to 2056, which indicates that the proposed algorithm can effectively reduce the number of trajectory interruptions and identity switching.

A laser driver which features high stability and a graphical user-interface was designed and used in trace gas sensing. The running of the laser driver was managed by an ARM processor which was embedded with a real-time operating system (RTOS). Through clicking on the touch screen that was configured with an emWin graphical user-interface (GUI), the parameters of the driving current can be graphically set and monitored. The circuit model of the distributed feedback laser diode (DFB-LD) was introduced into a TINA-SPICE simulation to evaluate the performance of the current source. Through simulation, the potential self-oscillation can be visually predicted, and the feedback loop can be appropriately compensated. To validate the applicability, the laser driver was used for driving a carefully selected DFB-LD and was employed in wavelength modulation spectroscopy (WMS) for CH4 detection at R(3) absorption line of the 2ν3 overtone. Under the conditions of room temperature, normal pressure and an effective absorption path of 15.4 cm, repetitive experiments were conducted using gas samples, with their concentrations ranging from 400 ppm to 1%, and the detection limit derived from the signal-to-noise ratio (SNR) was 7.2 ppm. The promising result indicates the high potential of this laser driver for use in absorption spectrum-based sensing applications.

The largest number of cybersecurity attacks is on web applications, in which Cross-Site Scripting (XSS) is the most popular way. The code audit is the main method to avoid the damage of XSS at the source code level. However, there are numerous limits implementing manual audits and rule-based audit tools. In the age of big data, it is a new research field to assist the manual auditing through machine learning. In this paper, we propose a new way to audit the XSS vulnerability in PHP source code snippets based on a PHP code parsing tool and the machine learning algorithm. We analyzed the operation sequence of source code and built a model to acquire the information that is most closely related to the XSS attack in the data stream. The method proposed can significantly improve the recall rate of vulnerability samples. Compared with related audit methods, our method has high reusability and excellent performance. Our classification model achieved an F1 score of 0.92, a recall rate of 0.98 (vulnerable sample), and an area under curve (AUC) of 0.97 on the test dataset.