View Synthesis with Scene Recognition for Cross-View Image Localization

Image-based localization has been widely used for autonomous vehicles, robotics, augmented reality, etc., and this is carried out by matching a query image taken from cell phone or vehicle dashcam to large scale of geo-tagged reference images, such as satellite/aerial images Google Street Views. However, the problem remains challenging due inconsistency between large-scale datasets regarding various light weather conditions. To tackle issue, work proposes novel view synthesis framework equipped with deep generative models, which can merge unique features outdated dataset containing seasonal changes. Our design scheme ensure that synthesized contain important both patch covering seasonable minimizing gap image-based tasks. The performance evaluation shows proposed synthesize views in lighting