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

This paper introduces an approach to produce accurate 3D detection boxes forobjects on the ground using single monocular images. We do so by merging 2Dvisual cues, 3D object dimensions, and ground plane constraints to produceboxes that are robust against small errors and incorrect predictions. First, wetrain a single-shot convolutional neural network (CNN) that produces multiplevisual and geometric cues of interest: 2D bounding boxes, 2D keypoints ofinterest, coarse object orientations and object dimensions. Subsets of thesecues are then used to poll probable ground planes from a pre-computed databaseof ground planes, to identify the "best fit" plane with highest consensus. Onceidentified, the "best fit" plane provides enough constraints to successfullyconstruct the desired 3D detection box, without directly predicting the 6DoFpose of the object. The entire ground plane polling (GPP) procedure isconstructed as a non-parametrized layer of the CNN that outputs the desired"best fit" plane and the corresponding 3D keypoints, which together define thefinal 3D bounding box. Doing so allows us to poll thousands of different groundplane configurations without adding considerable overhead, while also creatinga single CNN that directly produces the desired output without the need forpost processing. We evaluate our method on the 2D detection and orientationestimation benchmark from the challenging KITTI dataset, and provide additionalcomparisons for 3D metrics of importance. This single-stage, single-pass CNNresults in superior localization and orientation estimation compared to morecomplex and computationally expensive monocular approaches.

Ground Plane Polling for 6DoF Pose Estimation of Objects on the Road