TY - GEN
T1 - Bringing 3D Models Together
T2 - 13th Asian Conference on Computer Vision, ACCV 2016
AU - Wang, Ke
AU - Dunn, Enrique
AU - Rodriguez, Mikel
AU - Frahm, Jan Michael
N1 - Publisher Copyright:
© Springer International Publishing AG 2017.
PY - 2017
Y1 - 2017
N2 - The recent advances in large-scale scene modeling have enabled the automatic 3D reconstruction of landmark sites from crowdsourced photo collections. Here, we address the challenge of leveraging crowdsourced video collections to identify connecting visual observations that enable the alignment and subsequent aggregation, of disjoint 3D models. We denote these connecting image sequences as video liaisons and develop a data-driven framework for fully unsupervised extraction and exploitation. Towards this end, we represent video contents in terms of a histogram representation of iconic imagery contained within existing 3D models attained from a photo collection. We then use this representation to efficiently identify and prioritize the analysis of individual videos within a large-scale video collection, in an effort to determine camera motion trajectories connecting different landmarks. Results on crowdsourced data illustrate the efficiency and effectiveness of our proposed approach.
AB - The recent advances in large-scale scene modeling have enabled the automatic 3D reconstruction of landmark sites from crowdsourced photo collections. Here, we address the challenge of leveraging crowdsourced video collections to identify connecting visual observations that enable the alignment and subsequent aggregation, of disjoint 3D models. We denote these connecting image sequences as video liaisons and develop a data-driven framework for fully unsupervised extraction and exploitation. Towards this end, we represent video contents in terms of a histogram representation of iconic imagery contained within existing 3D models attained from a photo collection. We then use this representation to efficiently identify and prioritize the analysis of individual videos within a large-scale video collection, in an effort to determine camera motion trajectories connecting different landmarks. Results on crowdsourced data illustrate the efficiency and effectiveness of our proposed approach.
UR - https://www.scopus.com/pages/publications/105036995537
UR - https://www.scopus.com/pages/publications/105036995537#tab=citedBy
U2 - 10.1007/978-3-319-54190-7 25
DO - 10.1007/978-3-319-54190-7 25
M3 - Conference contribution
AN - SCOPUS:105036995537
SN - 9783319541891
T3 - Lecture Notes in Computer Science
SP - 408
EP - 423
BT - Computer Vision – ACCV 2016 - 13th Asian Conference on Computer Vision, Revised Selected Papers, Part 4
A2 - Lai, Shang-Hong
A2 - Nishino, Ko
A2 - Lepetit, Vincent
A2 - Sato, Yoichi
Y2 - 20 November 2016 through 24 November 2016
ER -