This Special Issue reprint provides an overview of object detection in images and videos, with a focus on addressing the resource constraints of lightweight vision sensors. Object detection has long been a key research area in computer vision. It is now gaining increasing attention from both academia and industry, driven by the rapid advancement of deep neural networks (DNNs) and high-resolution vision sensors. While DNNs have achieved remarkable success in recent years, they are becoming increasingly complex, with deeper network structures and larger training datasets. This growing complexity poses a challenge for deploying computationally and data-intensive DNNs on resource-limited vision sensors, particularly for real-time object detection.