Huang, Ke TingKe TingHuangSun, Shih WeiShih WeiSunHu, Min ChunMin ChunHuWEN-HUANG CHENGTsai, Tsung HungTsung HungTsai2023-03-132023-03-132012-12-319781450319119https://scholars.lib.ntu.edu.tw/handle/123456789/629227Digital photos are estimated to be generated and uploaded to social image websites (e.g. Flickr) in the amount of 375 petabytes annually [Tueld et al.], and plenty of these photos are annotated with the creation time/date in EXIF information. Among these photos, a large number of images are geo-tagged (e.g. GPS and landmark names) to indicate where the photo was taken. Moreover, several public databases composed of pictures captured by live cameras at various locations are available on the internet . Therefore, it has become a crucial task to enable an intuitive and efficient way for users to browse large-scale photos in space and time. Conventional photo browsing applications, e.g. ACDSee and Picasa, display photos one by one according to alphabetical or temporal order. To enrich the user's browsing experience, numerous multimedia systems, such as Photo Story, Tiling Slideshow, Auto- Collage, and IS-Scrapbook, have been developed to organize photos based on various photo metadata, including time information, visual contents, and typographical design principles. Aiming at creating a browsing style with a realistic feel, Photo Navigator is proposed to reveal the spatial relations among photos and offer a strong sense of space by taking users to fly into the scenes [Hsieh et al.]. Instead of regarding the system users as passive content receivers, Google Street View and Photo Tourism [Snavely et al.] allow the users to interact with the systems and actively determine the browsing path based on the built spatial relations among photos by the techniques of image stitching and 3D reconstruction. However, the temporal relations of photos are not well taken into account. Copyright is held by the author / owner(s).conference paper10.1145/2407156.24071812-s2.0-84871579693https://api.elsevier.com/content/abstract/scopus_id/84871579693