Deep learning-based vehicle event identification

In recent years, there has been a rapid development of intelligent driving assistance systems. Although most vehicles nowadays are equipped with driving assistance systems, the number of car accidents continues to rise. The main cause of car accidents is still largely attributed to human factors. Therefore, there has been an increasing focus on research related to vehicle event recognition and traffic video analysis. This study used deep learning methods to automatically recognize vehicle events from recorded driving videos. We expanded Inception-V3 to a 3D structure and replaced the bottom architecture of SlowFastNet with 3D-Inception-V3 to form a 3D-Inception-V3 + SlowFastNet model, making the network more suitable for video data recognition. In the experiment, we cut all videos into one-second clips and label each clip with vehicle event categories, including vehicle stopping, straight driving, turning, and collision. The experiment utilized training and testing procedures with eight distinct models, i.e., R3D, I3D, I3D + Kinetics-400, SlowFastNet(50), SlowFastNet(101), SlowFastNet(152), SlowFastNet(200), and the proposed 3D-Inception-V3 + SlowFastNet models. Comparative analysis through experiments revealed that our proposed network achieved the highest recognition accuracy of 93.3%.