Recovering the Spatial Layout of
Cluttered Rooms
Abstract:
In this
paper, we consider the problem of recovering the spatial layout of
indoor scenes from monocular images. The presence of clutter is a major
problem for existing singleview 3D reconstruction algorithms, most of
which rely on finding the ground-wall boundary. In most rooms, this
boundary is partially or entirely occluded. We gain robustness to
clutter by modeling the global room space with a parameteric 3D “box”
and by iteratively localizing clutter and refitting the box. To fit the
box, we introduce a structured learning algorithm that chooses the set
of parameters to minimize error, based on global perspective cues. On a
dataset of 308 images, we demonstrate the ability of our algorithm to
recover spatial layout in cluttered rooms and show several examples of
estimated free space.
|
|
Varsha Hedau, Derek Hoiem, David Forsyth, “Recovering the Spatial
Layout of Cluttered Rooms,” in the Twelfth IEEE International
Conference on Computer Vision, 2009.
@inproceedings{Hedau_2009_SpatialLayout, Author = {Varsha Hedau and Derek Hoiem and David Forsyth}, Title = {Recovering the Spatial Layout of Cluttered Rooms}, Booktitle = {Proceedings of the {IEEE} International Conference on Computer Vision ({ICCV '09})}, Year = {2009}, }
Data:
Groundtruth data
used in the above work. (314 Images 12 MB).
ReadMe.txt
To obtain the code below, please email me at
Executable:
Computation
of triplet of orthogonal vanishing points in an image, given detected
straight line segments. This code is distributed as an executable
compiled using matlab compiler. [Code] [ReadMe.txt] [MCRInstaller]
Spatial Layout Code:
Test
code for estimating spatial layout of an image as described in ICCV'09
paper. Feature computation uses integral images in
rectified domain, as described in ECCV'10. This is mainly done for
speedup and might result into small differences in results. This
package also contains source code for vanishing points executable
provided above. This code is distributed as is for research purpose only. [Code] [ReadMe.txt]
Demo code for free space estimation method described in ICCV'09 paper. [Code]
|