SMASH: Physics-guided Reconstruction of Collisions from Videos

03/29/2016
by   Aron Monszpart, et al.
0

Collision sequences are commonly used in games and entertainment to add drama and excitement. Authoring even two body collisions in the real world can be difficult, as one has to get timing and the object trajectories to be correctly synchronized. After tedious trial-and-error iterations, when objects can actually be made to collide, then they are difficult to capture in 3D. In contrast, synthetically generating plausible collisions is difficult as it requires adjusting different collision parameters (e.g., object mass ratio, coefficient of restitution, etc.) and appropriate initial parameters. We present SMASH to directly read off appropriate collision parameters directly from raw input video recordings. Technically we enable this by utilizing laws of rigid body collision to regularize the problem of lifting 2D trajectories to a physically valid 3D reconstruction of the collision. The reconstructed sequences can then be modified and combined to easily author novel and plausible collisions. We evaluate our system on a range of synthetic scenes and demonstrate the effectiveness of our method by accurately reconstructing several complex real world collision events.

READ FULL TEXT

page 1

page 2

page 6

page 7

page 8

research
03/25/2023

A novel collision model for inextensible textiles and its experimental validation

In this work, we introduce a collision model specifically tailored for t...
research
01/21/2021

COLLIDE-PRED: Prediction of On-Road Collision From Surveillance Videos

Predicting on-road abnormalities such as road accidents or traffic viola...
research
07/23/2020

High Precision Real Time Collision Detection

Collision detection and collision avoidance are essential components in ...
research
04/18/2023

CabiNet: Scaling Neural Collision Detection for Object Rearrangement with Procedural Scene Generation

We address the important problem of generalizing robotic rearrangement t...
research
08/25/2023

Small Celestial Body Exploration with CubeSat Swarms

This work presents a large-scale simulation study investigating the depl...
research
04/15/2019

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

We introduce an approach to model surface properties governing bounces i...
research
06/19/2018

Single-View Hair Reconstruction using Convolutional Neural Networks

We introduce a deep learning-based method to generate full 3D hair geome...

Please sign up or login with your details

Forgot password? Click here to reset