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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Object localization in 3D space is a challenging aspect in monocular 3Dobject detection. Recent advances in 6DoF pose estimation have shown thatpredicting dense 2D-3D correspondence maps between image and object 3D modeland then estimating object pose via Perspective-n-Point (PnP) algorithm canachieve remarkable localization accuracy. Yet these methods rely on trainingwith ground truth of object geometry, which is difficult to acquire in realoutdoor scenes. To address this issue, we propose MonoRUn, a novel detectionframework that learns dense correspondences and geometry in a self-supervisedmanner, with simple 3D bounding box annotations. To regress the pixel-related3D object coordinates, we employ a regional reconstruction network withuncertainty awareness. For self-supervised training, the predicted 3Dcoordinates are projected back to the image plane. A Robust KL loss is proposedto minimize the uncertainty-weighted reprojection error. During testing phase,we exploit the network uncertainty by propagating it through all downstreammodules. More specifically, the uncertainty-driven PnP algorithm is leveragedto estimate object pose and its covariance. Extensive experiments demonstratethat our proposed approach outperforms current state-of-the-art methods onKITTI benchmark.

MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty Propagation