Technical Papers

Acquiring and Synthesizing Indoor Scenes

Thursday, 29 November 14:15 - 16:00 |  Peridot 206

Example-based Synthesis of 3D Object Arrangements - Picture

Example-based Synthesis of 3D Object Arrangements

We present a method for synthesizing 3D object arrangements from examples. Given a few user-provided examples, our system can synthesize a diverse set of plausible new scenes by learning from a larger scene database.


Matthew Fisher, Stanford University
Daniel Ritchie, Stanford University
Manolis Savva, Stanford University
Thomas Funkhouser, Princeton University
Pat Hanrahan, Stanford University


An Interactive Approach to Semantic Modeling of Indoor Scenes with an RGBD Camera - Picture

An Interactive Approach to Semantic Modeling of Indoor Scenes with an RGBD Camera

An approach to enable you to create 3D model of your room with Kinect camera in a few minutes


Tianjia Shao, Tsinghua University
Weiwei Xu, Hangzhou Normal University
Kun Zhou, Zhejiang University
Jingdong Wang, Microsoft Research Asia
Dongping Li, Zhejiang University
Baining Guo, Microsoft Research Asia


A Search-Classify Approach for Cluttered Indoor Scene Understanding - Picture

A Search-Classify Approach for Cluttered Indoor Scene Understanding

We present a search-classify approach for recognition and reconstruction of scanned 3D indoor scenes. Our method interleaves between segmentation and classification in an iterative manner. We reinforce classification by a template fitting step which also yields scene approximation.


Liangliang Nan, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences
Ke Xie, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences
Andrei Sharf, Ben-Gurion University, Israel


Acquiring 3D Indoor Environments with Variability and Repetition - Picture

Acquiring 3D Indoor Environments with Variability and Repetition

We introduce a pipeline to (1) acquire
3D models of common furniture consisting of rigid
parts, as well as of their low-dimensional variability modes; (2)
detect and recognize occurrences of such models from single low
quality scans; and (3) quickly populate large indoor environments
enabling novel reconstruction possibilities.


Young Min Kim, Stanford University
Niloy Mitra, University College London
Dongming Yan, King Abdullah University of Science and Technology
Leonidas Guibas, Stanford University