cyberagent/in-store-visual-localization

Dataset for monocular visual relocalization on a COLMAP 3D reconstruction model. This dataset was collected at EZOHUB Tokyo with the cooperation of SATUDORA HOLDINGS CO.,LTD. Dataset structure train : Images used to construct the COLMAP model. model : The COLMAP model files. test : Test images to run relocalization. intrinsic.xml : Intrinsic camera parameters.