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

We introduce an approach to model surface properties governing bounces ineveryday scenes. Our model learns end-to-end, starting from sensor inputs, topredict post-bounce trajectories and infer two underlying physical propertiesthat govern bouncing - restitution and effective collision normals. Our model,Bounce and Learn, comprises two modules -- a Physics Inference Module (PIM) anda Visual Inference Module (VIM). VIM learns to infer physical parameters forlocations in a scene given a single still image, while PIM learns to modelphysical interactions for the prediction task given physical parameters andobserved pre-collision 3D trajectories. To achieve our results, we introducethe Bounce Dataset comprising 5K RGB-D videos of bouncing trajectories of afoam ball to probe surfaces of varying shapes and materials in everyday scenesincluding homes and offices. Our proposed model learns from our collecteddataset of real-world bounces and is bootstrapped with additional informationfrom simple physics simulations. We show on our newly collected dataset thatour model out-performs baselines, including trajectory fitting with Newtonianphysics, in predicting post-bounce trajectories and inferring physicalproperties of a scene.

Bounce and Learn: Modeling Scene Dynamics with Real-World Bounces