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

Current RGB-based 6D object pose estimation methods have achieved noticeableperformance on datasets and real world applications. However, predicting 6Dpose from single 2D image features is susceptible to disturbance from changingof environment and textureless or resemblant object surfaces. Hence, RGB-basedmethods generally achieve less competitive results than RGBD-based methods,which deploy both image features and 3D structure features. To narrow down thisperformance gap, this paper proposes a framework for 6D object pose estimationthat learns implicit 3D information from 2 RGB images. Combining the learned 3Dinformation and 2D image features, we establish more stable correspondencebetween the scene and the object models. To seek for the methods best utilizing3D information from RGB inputs, we conduct an investigation on three differentapproaches, including Early- Fusion, Mid-Fusion, and Late-Fusion. We ascertainthe Mid- Fusion approach is the best approach to restore the most precise 3Dkeypoints useful for object pose estimation. The experiments show that ourmethod outperforms state-of-the-art RGB-based methods, and achieves comparableresults with RGBD-based methods.

Towards Two-view 6D Object Pose Estimation: A Comparative Study on Fusion Strategy