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

A major challenge in monocular 3D object detection is the limited diversityand quantity of objects in real datasets. While augmenting real scenes withvirtual objects holds promise to improve both the diversity and quantity of theobjects, it remains elusive due to the lack of an effective 3D object insertionmethod in complex real captured scenes. In this work, we study augmentingcomplex real indoor scenes with virtual objects for monocular 3D objectdetection. The main challenge is to automatically identify plausible physicalproperties for virtual assets (e.g., locations, appearances, sizes, etc.) incluttered real scenes. To address this challenge, we propose a physicallyplausible indoor 3D object insertion approach to automatically copy virtualobjects and paste them into real scenes. The resulting objects in scenes have3D bounding boxes with plausible physical locations and appearances. Inparticular, our method first identifies physically feasible locations and posesfor the inserted objects to prevent collisions with the existing room layout.Subsequently, it estimates spatially-varying illumination for the insertionlocation, enabling the immersive blending of the virtual objects into theoriginal scene with plausible appearances and cast shadows. We show that ouraugmentation method significantly improves existing monocular 3D object modelsand achieves state-of-the-art performance. For the first time, we demonstratethat a physically plausible 3D object insertion, serving as a generative dataaugmentation technique, can lead to significant improvements for discriminativedownstream tasks such as monocular 3D object detection. Project website:https://gyhandy.github.io/3D-Copy-Paste/

3D Copy-Paste: Physically Plausible Object Insertion for Monocular 3D Detection