Driver Action Recognition in Low-Light Conditions: A Multi-View Fusion Framework
Karam Abdullah, Imen Jegham, Mohamed Ali Mahjoub, Anouar Ben Khalifa · 2024
As big data continues to rise, the fusion of information emerges as one of the primary concerns in the fields of computer vision and intelligent transportation systems. A multi-view driver action recognition system can be a key to road safety, especially at nighttime. In fact, driver monitoring from a single view is one of the most challenging task due to the high amount of issues present including occlusion and illumination variation. A multiview driver action recognition system can provide more abundant and complementary driver behavior information which helps to prevent dramatic traffic injuries. In this work, we introduce a multi-viewpoint information fusion system that reliably classifies drivers’ actions at nighttime. Experiments on the public and unique realistic dataset, 3MDAD, demonstrate that our proposed approach achieves high recognition accuracy compared to state-of-the-art methods.