2martens/masterthesis-latex
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Added paragraph about adapting GPND to SceneNet and SSD

Browse Source 
Signed-off-by: Jim Martens <github@2martens.de>
This commit is contained in:
Jim Martens
2019-03-07 18:14:39 +01:00
parent b52391ded3
commit 66f81284fb
1 changed files with 34 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
body_expose.tex
View File

@ -331,6 +331,35 @@ Practically, the auto-encoder is trained separately for every
object class that is considered "known". Pidhorskyi et al trained object class that is considered "known". Pidhorskyi et al trained
it on the MNIST\cite{Lecun1998} data set, once for every digit. it on the MNIST\cite{Lecun1998} data set, once for every digit.
For this thesis it needs to be trained on the SceneNet RGB-D
data set using MS COCO classes as known classes. As in every
test epoch all known classes are present, it becomes
non-trivial which of the trained auto-encoders should be used to
calculate novelty. To phrase it differently, a true positive
detection is possible for multiple classes in the same image.
If, for example, one object is classified correctly by SSD as a chair
the novelty score should be low. But the auto-encoders of all
known classes but the "chair" class will give ideally a high novelty
score. Which of the values should be used? The only sensible solution
is to only run it through the auto-encoder that was trained for
the class the SSD model predicted. This provides the following
scenarios:
\begin{itemize}
\item true positive classification: novelty score should be low
\item false positive classification and correct class is
among the known classes: novelty score should be high
\item false positive classification and correct class is unknown:
novelty score should be high
\end{itemize}
\noindent
Negative classifications are not listed as these are not part
of the output of the SSD and cannot be given to the auto-encoder
as input. Furthermore, the 2nd case should not happen because
the trained SSD knows this other class and is very likely
to give it a higher probability. Therefore, using only one
auto-encoder fulfils the task of differentiating between
known and unknown classes.
\section{Contribution} \section{Contribution}
This section will outline what exactly the scientific as well as This section will outline what exactly the scientific as well as
@ -356,11 +385,11 @@ a better object detection performance. The GPND approach uses
the auto-encoder losses and results to identify novel cases and the auto-encoder losses and results to identify novel cases and
therefore mark detections as false positive. Subsequently these therefore mark detections as false positive. Subsequently these
detections can be discarded as well. By comparing the object detections can be discarded as well. By comparing the object
detection performance after discarding the identified false detection performance after discarding the identified false positive
positive cases, the effectiveness of both approaches can be cases, the effectiveness of both approaches can be compared with each
compared with each other. It is interesting to research if the other. It is interesting to research if the GPND approach results in
GPND approach results in a better object detection performance a better object detection performance than the dropout sampling
than the dropout sampling provides. provides.
The formulated hypothesis, which is repeated after this paragraph, The formulated hypothesis, which is repeated after this paragraph,
combines both aspects and requires a similar or better result in combines both aspects and requires a similar or better result in