Conference 5–9 August 2012
Exhibition 7–9 August 2012
Los Angeles Convention Center

All About Images

Technical Papers

All About Images

Thursday, 9 August 9:00 AM - 10:30 AM | Los Angeles Convention Center, Room 515AB
Session Chair: James Hays, Brown University

Interactive Images: Cuboid Proxies for Smart Image Manipulation

Interactive images create a partial scene reconstruction of a single image based on cuboid-proxies with minimal user interaction and subsequently allow believable scene-level edits while conforming to extracted non-local relations.

Youyl Zheng
King Abdullah University of Science and Technology

Xiang Chen
Zhejiang University

Ming-Ming Cheng
Tsinghua University

Kun Zhou
Zhejiang University

Shi-Min Hu
Tsinghua University

Niloy Mitra
University College London

A Framework for Content-Adaptive Photo Manipulation Macros: Application to Face, Landscape, and Global Manipulations

This framework for generating content-adaptive macros that can transfer complex photo manipulations to new target images uses automated image labeling and machine-learning techniques to learn the dependencies between image features and the parameters of each selection, brush stroke, and image-processing operation in the macro.

Floraine Berthouzoz
University of California, Berkeley

Wilmot Li
Adobe Systems Incorporated

Mira Dontcheva
Adobe Systems Incorporated

Maneesh Agrawala
University of California, Berkeley

Image-Based Rendering for Scenes With Reflections

This paper describes an image-based rendering system that uses two additive layers, each with its own proxy geometry, to model reflections in real-world scenes. It shows how to recover such descriptions from real-world image sequences and how to re-render them in a realistic fashion.

Sudipta N. Sinha
Microsoft Research

Johannes Kopf
Microsoft Research

Michael Goesele
Technische Universität Darmstadt

Daniel Scharstein
Middlebury College

Richard Szeliski
Microsoft Research

What Makes Paris Look Like Paris?

Given a large repository of geotagged imagery provided by Google Street View, this paper shows how to automatically find local visual elements, such as street signs, windows, or balconies, that are most distinctive for a given locale, such as the city of Paris.

Carl Doersch
Carnegie Mellon University

Saurabh Singh
Carnegie Mellon University

Abhinav Gupta
Carnegie Mellon University

Josef Sivic
INRIA

Alexei A. Efros
Carnegie Mellon University