Back to search

Cornell University

Applications of numerical optimization in graphics and games

Abstract

dc:description.abstract

This dissertation begins with a camera calibration system, Caliber, which solves pose estimation problems consisting of two types of constraints: relative pose constraints, resulting from measurements, such as found in SLAM and motion estimation problems; and rigidity constraints, the notion of objects that are rigidly attached to each other so that their relative pose is fixed over time even if that pose is not known a priori. We show that this problem is NP-hard, but demonstrate an algorithm that works well in practice. Applications include calibrating goniometers for accurate measurement of light reflection. We then proceed further down the pipeline from measurement of light reflection to the modeling of the appearance of surfaces. Specifically, we examine the appearance of wood. While suitable BRDF models exist, the texture parameter maps for these wood BRDFs are difficult to author---good results have been shown with elaborate measurements for small flat samples, but these models are not much used in practice. Furthermore, mapping 2D image textures onto 3D objects leads to distortion and inconsistencies. Procedural volumetric textures solve these geometric problems, but existing methods produce much lower quality than image textures. This chapter aims to bring the best of all these techniques together: we present a comprehensive volumetric simulation of wood appearance, including growth rings, color variation, pores, rays, and growth distortions. The fiber directions required for anisotropic specular figure follow naturally from the distortions. The final piece of the dissertation is another application of numerical optimization, this time in games, both in the sense of entertainment and in the sense of game theory. A key challenge in game design is achieving balance between the strategies available to the players. We model the balancing problem as modifying a zero-sum game, using one variable per strategy, so that every strategy has an incentive to be employed. We begin with a special case where these variables affect player payoffs multiplicatively, and show that the simple Sinkhorn-Knopp algorithm can be used to balance the game. We then proceed to analyze the more general case where the variables have a monotonic effect on payoffs, and show that it is amenable to standard optimization methods.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Computer Science
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Computer Science
Grantor
Cornell University
Year dc:date.issued
2018

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Liu, Albert Julius
Chair dc:contributor.chair
  • Marschner, Stephen Robert
Committee members dc:contributor.committeemember
  • Vladimirsky, Alexander B.
  • Kleinberg, Robert David
  • Snavely, Keith Noah

Subjects

dc:subject × 6

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Dc Identifier Bibid
10489513
Dc Identifier Other
ProQuest Submission ID: 10724
ProQuest Publication ID: 10745566

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Liu, Albert Julius. Applications of numerical optimization in graphics and games. Doctor of Philosophy thesis, Cornell University, 2018. https://hdl.handle.net/1813/59428