2martens
/
uni
mirror of https://github.com/2martens/uni.git
1
0
Fork 0
Code Releases Activity

[Masterproj] Added section about experiments 

Signed-off-by: Jim Martens <github@2martens.de>
This commit is contained in:
Jim Martens 2018-05-22 14:08:40 +02:00
parent d062527354
commit d4de03f9d6
2 changed files with 108 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
masterproj/bib.bib
View File

@ -13,6 +13,30 @@
Timestamp = {2018.05.08}
}
@Inproceedings{Gupta2015,
Title = {Aligning 3D models to RGB-D images of cluttered scenes},
Author = {Gupta, Saurabh and Arbel{\'a}ez, Pablo and Girshick, Ross and Malik, Jitendra},
Booktitle = {Computer Vision and Pattern Recognition (CVPR), 2015 IEEE Conference on},
Year = {2015},
Pages = {4731--4740},
Publisher = {IEEE},
Owner = {jim},
Timestamp = {2018.05.22}
}
@Inproceedings{Silberman2012,
Title = {Indoor segmentation and support inference from RGBD images},
Author = {Silberman, Nathan and Hoiem, Derek and Kohli, Pushmeet and Fergus, Rob},
Booktitle = {European Conference on Computer Vision},
Year = {2012},
Pages = {746--760},
Publisher = {Springer},
Owner = {jim},
Timestamp = {2018.05.22}
}
@Inproceedings{Simonyan2015,
Title = {Very Deep Convolutional Networks for Large-Scale Image Recognition},
Author = {Karen Simonyan and Andrew Zisserman},
@ -23,6 +47,18 @@
Timestamp = {2018.05.02}
}
@Inproceedings{Song2015,
Title = {SUN RGB-D: A RGB-D scene understanding benchmark suite},
Author = {Song, Shuran and Lichtenberg, Samuel P and Xiao, Jianxiong},
Booktitle = {CVPR},
Year = {2015},
Pages = {567--576},
Publisher = {IEEE},
Owner = {jim},
Timestamp = {2018.05.22}
}
@Inproceedings{Song2016,
Title = {Deep Sliding Shapes for Amodal 3D Object Detection in RGB-D Images},
Author = {S. Song and J. Xiao},
@ -34,3 +70,15 @@
Timestamp = {2018.05.02}
}
@Inproceedings{Song2014,
Title = {Sliding Shapes for 3D object detection in depth images},
Author = {Song, Shuran and Xiao, Jianxiong},
Booktitle = {European conference on computer vision},
Year = {2014},
Pages = {634--651},
Publisher = {Springer},
Owner = {jim},
Timestamp = {2018.05.22}
}

62
masterproj/seminar_report.tex
View File

@ -255,8 +255,66 @@ scores for every box. In case of the box regressions the results from the networ
are used directly.
\section{Experimental result and evaluation}
In this section, the evaluation and experimental results of proposed method should be described.
Also provide some discussion, answering questions such as: when does the method work well, when not? How does it compare to other state-of-the-art works?
The regional proposal network was trained for 10 hours and the object recognition
network was trained for 17 hours. In both cases an Nvidia K40 GPU was used.
During testing phase it took the RPN \(5.62\) seconds per image and the ORN
\(13.93\) seconds per image. Both networks were evaluated on the NYUv2\cite{Silberman2012}
and SUN RGB-D\cite{Song2015} data sets.
A threshold of \(0.25\) was used to calculate the average recall for the proposal
generation and the average precision for the detection. The SUN RGB-D data set
was used to obtain the ground truth amodal bounding boxes.
For the evaluation of the proposal generation a single-scale RPN, a multi-scale RPN
and a multi-scale RPN with RGB colour added to the 3D TSDF were compared with
each other and the baselines using the NYU data set. 3D selective search
and a naive 2D to 3D conversion were used as baselines. The naive conversion used the
2D region proposal to retrieve the 3D points within that region. Afterwards the
outermost 2 percentiles in each direction were removed and a tight 3D bounding
box calculated. The values of recall averaged over all object categories were
\(34.4\) for the naive approach, \(74.2\) for 3D selective search, \(75.2\) for
the single-scale RPN, \(84.4\) for the multi-scale RPN and \(84.9\) for the
multi-scale RPN with added colour. The last value is used as the final region
proposal result.
Another experiment tested the detection results for the same ORN architecture
given different region proposals. Comparing the 3D selective search with
RPN gave mean average precisions of \(27.4\) and \(32.3\) respectively. Hence
the RPN provides a better solution. Planar objects (e.g. doors) seem to work
better with 3D selective search. Boxes, monitors and TVs don't work for the RPN,
where the presumed reason for boxes is the high variance and for monitors and TVs
the missing depth information is likely responsible.
The detection evaluation was structured differently. First the feature encodings
were compared with other (the same experiment that was mentioned in previous
paragraph), then the design was justified and lastly the results were compared
with state-of-the-art methods. The feature encoding experiment provided better
results for encoding the directions directly compared to a single distance.
An accurate TSDF measured better than a projective one. The usage of the 2D image
VGGnet proved to be better than the direct encoding of colour on 3D voxels.
Lastly it didn't help to include HHA (horizontal disparity, height above ground
and the angle the pixel's local surface normal makes with the inferred gravity
direction).
The same experiment was used to help with design choices. It was found that
bounding box regression helps significantly (increase in mAP of 4.4 and 4.1
for 3D selective search and RPN respectively compared to the case without this
regression). SVM was found to outperform the softmax slightly (increase of 0.5 mAP)
which presumably is the case, because it can better handle the unbalanced number of
training samples for each category in the NYUv2 data set. Size pruning was identified
as helping (increase of mAP per category of 0.1 up to 7.8).
For the comparison with state-of-the-art methods Song and Xiao used 3D Sliding
Shapes\cite{Song2014} and 2D Depth-RCNN\cite{Gupta2015} and the same test set
that was used for the 2D Depth-RCNN (intersection of NYUv2 test set and Sliding
Shapes test set for the five categories bed, chair, table, sofa/couch and toilet).
The comparison shows that 3D Deep Sliding Shapes outperforms the chosen state-of-the-art
methods in all categories. The toilet is the only example where it is relevant
for the result that the 2D data is used. With only 3D data used the 2D Depth-RCNN
performs better on the estimated model if it uses 2D and 3D.
All in all 3D Deep Sliding Shapes works well on non-planar objects that have depth
information. The 2D component helps in distinguishing similar shaped objects.
\section{Discussion} % (fold)
\label{sec:discussion}