Honors College Thesis
 

Solving Physical Reasoning Tasks in Simulated Environments

Pubblico Deposited

Contenuto scaricabile

Scarica il pdf
https://ir.library.oregonstate.edu/concern/honors_college_theses/xd07h140g

Descriptions

Attribute NameValues
Creator
Abstract
  • We take for granted how quickly we, as humans, form mental models of the world around us. By the time we are toddlers, we have an observable intuition around the physical rules of the world. Stacking blocks such that they don’t fall over becomes such a trivial task, that it enters the sphere of common sense. We aim to develop an agent with such physical common sense capabilities inspired by the train of thought that humans build mental models and simulations of their environment in real time. Our agent forms a hypothesis for the dynamics of objects in a given scene through the reconstruction and subsequent “mental” simulation of observed objects. This hypothesis comes in the form of both a quantitative comparison of the scene to the hypothesis, and a qualitative assessment of the scene’s compliance with expectations. Our agent was evaluated on its ability to solve a gravity based task, and received top marks among competing agents.
  • Key Words: machine common sense, computer vision, physical reasoning, physics engines
Contributor
Resource Type
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Non-Academic Affiliation
Dichiarazione dei diritti
Publisher
Peer Reviewed
Language

Le relazioni

Parents:

This work has no parents.

In Collection:

Elementi

Thumbnail Titolo Data caricata Visibilità Azioni
file details ShureihZeyad2021.pdf 2021-06-04 Pubblico Scaricare