|
Martin Čadík, cadikm@centrum.cz Jan Vašíček, xvasic21@stud.fit.vutbr.cz Michal Hradiš, ihradis@fit.vutbr.cz Filip Radenović, filip.radenovic@cmp.felk.cvut.cz Ondřej Chum, chum@cmp.felk.cvut.cz |
This work addresses the problem of camera elevation estimation from a single photograph in an outdoor environment. We introduce a new benchmark dataset of one-hundred thousand images with annotated camera elevation called Alps100K. We propose and experimentally evaluate two automatic data-driven approaches to camera elevation estimation: one based on convolutional neural networks, the other on local features. To compare the proposed methods to human performance, an experiment with 100 subjects is conducted. The experimental results show that both proposed approaches outperform humans and that the best result is achieved by their combination.
The dataset Alps100K has been collected from publicly available source flickr.com for non-commercial research and academic purposes. The rights, listed in file Alps100K_licenses.txt, to the photographs remain with the authors. Use of these images must respect the corresponing terms of use. Whenever you use the dataset, please acknowledge it by citing our BMVC 2015 paper.