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

Object pose recovery has gained increasing attention in the computer visionfield as it has become an important problem in rapidly evolving technologicalareas related to autonomous driving, robotics, and augmented reality. Existingreview-related studies have addressed the problem at visual level in 2D, goingthrough the methods which produce 2D bounding boxes of objects of interest inRGB images. The 2D search space is enlarged either using the geometryinformation available in the 3D space along with RGB (Mono/Stereo) images, orutilizing depth data from LIDAR sensors and/or RGB-D cameras. 3D bounding boxdetectors, producing category-level amodal 3D bounding boxes, are evaluated ongravity aligned images, while full 6D object pose estimators are mostly testedat instance-level on the images where the alignment constraint is removed.Recently, 6D object pose estimation is tackled at the level of categories. Inthis paper, we present the first comprehensive and most recent review of themethods on object pose recovery, from 3D bounding box detectors to full 6D poseestimators. The methods mathematically model the problem as a classification,regression, classification & regression, template matching, and point-pairfeature matching task. Based on this, a mathematical-model-based categorizationof the methods is established. Datasets used for evaluating the methods areinvestigated with respect to the challenges, and evaluation metrics arestudied. Quantitative results of experiments in the literature are analyzed toshow which category of methods best performs across what types of challenges.The analyses are further extended comparing two methods, which are our ownimplementations, so that the outcomes from the public results are furthersolidified. Current position of the field is summarized regarding object poserecovery, and possible research directions are identified.

A Review on Object Pose Recovery: from 3D Bounding Box Detectors to Full 6D Pose Estimators