Technical Papers

Image Enhancement

Thursday, 29 July | 3:45 PM - 5:15 PM | Room 403 AB
Session Chair: Dan Goldman, Adobe Systems Incorporated
Multi-Scale Image Harmonization

Compositing images that differ significantly in appearance often produces unrealistic results. This framework matches the visual appearance of images, including contrast, texture, noise, and blur, by manipulating their pyramid representations and blends them with alpha-based and seamless boundary constraints to produce highly realistic composites with minimal user interaction.

Kalyan Sunkavalli
Harvard University

Micah K. Johnson
Massachusetts Institute of Technology

Wojciech Matusik
Disney Research

Hanspeter Pfister
Harvard University

Personal Photo Enhancement Using Example Images

This framework for improving photos by using a person’s favorite photographs as examples aligns faces between “good” and “bad” photos and uses properties of the good examples to improve bad images. These “personal” photos provide strong guidance for a number of operations.

Neel Joshi
Microsoft Corporation

Wojciech Matusik
Disney Research

Edward H. Adelson
MIT CSAIL

David J. Kriegman
University of California, San Diego

Parametric Reshaping of Human Bodies in Images

An easy-to-use image retouching system that allows users to easily reshape a human body in a single image by simply manipulating a small set of sliders corresponding to semantic attributes such as height, weight, and waist girth.

Shizhe Zhou
Zhejiang University

Hongbo Fu
City University of Hong Kong

Ligang Liu
Zhejiang University

Daniel Cohen-Or
Tel-Aviv University

Xiaoguang Han
Zhejiang University

Image Warps for Artistic Perspective Manipulation

Painters commonly sketch vanishing points and lines to guide the construction of perspective images. This paper presents a tool for manipulating perspective in photographs via image-space controls similar to those used by artists. The system optimizes a warp that satisfies the user-specified constraints while minimizing distortion of the image.

Robert Carroll
University of California, Berkeley

Aseem Agarwala
Adobe Systems Incoporated

Maneesh Agrawala
University of California, Berkeley

In the News
SIGGRAPH 2010 Video