Technical Papers

Boundaries, Edges & Gradients

Wednesday, 28 July | 9:00 AM - 10:30 AM | Room 403 AB
Session Chair: Sylvain Paris, Adobe Systems Incorporated
RepFinder: Finding Approximately Repeated Scene Elements for Image Editing

A novel framework that uses rough inputs to detect and robustly extract repetitions along with their mutual relations. The paper demonstrates the versatility of the framework on a large set of inputs and demonstrates a number of applications.

Ming-ming Cheng
Tsinghua University

Fang-Lue Zhang
Tsinghua University

Niloy J. Mitra
Indian Institute of Technology Delhi, King Abdullah University of Science and Technology

Xiaolei Huang
Lehigh University

Shi-Min Hu
Tsinghua University

Edge-Based Image Coarsening

This new dimensionally reduced linear image space allows a number of recent image-manipulation techniques to be performed efficiently and robustly. Results show that the reduced representation is useful for energy-minimization methods in achieving efficient processing and providing better matrix conditioning with minimal quality sacrifice.

Raanan Fattal
Hebrew University

Robert Carroll
University of California, Berkeley

Maneesh Agrawala
University of California, Berkeley

GradientShop: A Gradient-Domain Optimization Framework for Image and Video Filtering

This paper presents an optimization framework for exploring gradient-domain solutions for image and video processing. The utility of the framework is demonstrated by creating several effective yet simple to implement solutions for various image-processing tasks.

Pravin Bhat
University of Washington

Distributed Gradient-Domain Processing of Planar and Spherical Images

This paper extends gradient-domain image processing by adapting the approach to operate on spherical imagery and proposing a distributed solver to efficiently process large images. The authors demonstrate the approach on several datasets, including a terapixel spherical scan of the night sky.

Michael Kazhdan
Johns Hopkins University

Dinoj Surendran
Microsoft Corporation

Hugues Hoppe
Microsoft Research

In the News
SIGGRAPH 2010 Video