Thursday, 28 July, 2:00 pm - 3:30 pm, Anaheim Convention Center, Ballroom C
Session Chair: Sylvain Lefebvre, Inria

Multi-Scale Label-Map Extraction for Texture Synthesis

This paper presents an unsupervised method for analyzing texture contents at multiple scales to support texture synthesis. It treats large inputs, and it can handle natural images containing different textured regions. The method is applied to non-stationary synthesis, content selection, and texture painting.

Yitzchak Lockerman
Yale University

Basile Sauvage
Université De Strasbourg

Remi Allegre
Université De Strasbourg

Jean-Michel Dischler
Université De Strasbourg

Julie Dorsey
Yale University

Holly Rushmeier
Yale University

Time-Varying Weathering in Texture Space

Given a single texture image, weathering degrees at different regions of the input texture are estimated by prevalence analysis of texture patches. This information then allows graceful increase or decrease of the popularity of weathered patches, simulating the evolution of texture appearance both backward and forward in time.

Rachele Bellini
Tel Aviv University

Yanir Kleiman
Tel Aviv University

Daniel Cohen-Or
Tel Aviv University

Vector-Regression Functions for Texture Compression

This paper proposes a representation (vector-regression functions) that uses neural networks to compactly approximate any point-sampled image and support GPU texture mapping, including random access and filtering operations.

Jiaping Wang

Ying Song
Zhejiang Sci-Tech University

Liyi Wei

Wencheng Wang
State Key Laboratory of Computer Science

Multiway K-Clustered Tensor Approximation: Toward High-Performance Photorealistic Data-Driven Rendering

This paper proposes a novel sparse multilinear model (MK-CTA) for rendering photorealistic images from large-scale multidimensional datasets. It can accurately and compactly represent complex datasets, while easily achieving high rendering rates.

Yu-Ting Tsai
Yuan Ze University