English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Though a large body of computer vision research has investigated developinggeneric semantic representations, efforts towards developing a similarrepresentation for 3D has been limited. In this paper, we learn a generic 3Drepresentation through solving a set of foundational proxy 3D tasks:object-centric camera pose estimation and wide baseline feature matching. Ourmethod is based upon the premise that by providing supervision over a set ofcarefully selected foundational tasks, generalization to novel tasks andabstraction capabilities can be achieved. We empirically show that the internalrepresentation of a multi-task ConvNet trained to solve the above core problemsgeneralizes to novel 3D tasks (e.g., scene layout estimation, object poseestimation, surface normal estimation) without the need for fine-tuning andshows traits of abstraction abilities (e.g., cross-modality pose estimation).In the context of the core supervised tasks, we demonstrate our representationachieves state-of-the-art wide baseline feature matching results withoutrequiring apriori rectification (unlike SIFT and the majority of learnedfeatures). We also show 6DOF camera pose estimation given a pair local imagepatches. The accuracy of both supervised tasks come comparable to humans.Finally, we contribute a large-scale dataset composed of object-centric streetview scenes along with point correspondences and camera pose information, andconclude with a discussion on the learned representation and open researchquestions.

Generic 3D Representation via Pose Estimation and Matching