The field of autonomous navigation and road safety is rapidly advancing with the development of innovative methods for camera pose estimation, road defect detection, and semantic segmentation. Recent research has focused on improving the accuracy and efficiency of Structure-from-Motion (SfM) systems, which are crucial for dense reconstruction in autonomous navigation and robotic perception. Additionally, there has been significant progress in the detection and management of urban defects, such as potholes and cracks, using computer vision and machine learning techniques. These advancements have the potential to improve road safety, reduce maintenance costs, and enhance the overall efficiency of transportation systems. Notable papers in this area include: CuSfM, which achieves significantly improved accuracy and processing speed compared to existing SfM methods. Semantic4Safety, which provides a framework for analyzing street-view imagery to derive insights into urban road safety. MRASfM, which enhances the reliability of camera pose estimation in driving scenes. InfraGPT Smart Infrastructure, which proposes a comprehensive pipeline for detecting and managing urban defects using street CCTV streams. StripRFNet, which achieves state-of-the-art accuracy and real-time efficiency in road damage detection. iWatchRoadv2, which presents a fully automated end-to-end platform for real-time pothole detection and geospatial mapping. Unsupervised Monocular Road Segmentation, which eliminates the reliance on costly manually labeled datasets for road segmentation. WP-CrackNet, which proposes a collaborative adversarial learning framework for end-to-end weakly-supervised road crack detection.