Unified entropy thresholds
Signed-off-by: Jim Martens <github@2martens.de>
This commit is contained in:
4
body.tex
4
body.tex
@ -134,7 +134,7 @@ baseline to compare against. In particular, vanilla SSD uses
|
||||
a per-class confidence threshold of 0.01, an IOU threshold of 0.45
|
||||
for the non-maximum suppression, and a top k value of 200.
|
||||
The effect of an entropy threshold is measured against this vanilla
|
||||
SSD by applying entropy thresholds from 0.1 to 2.4 (limits taken from
|
||||
SSD by applying entropy thresholds from 0.1 to 2.4 inclusive (limits taken from
|
||||
Miller et al.). Dropout sampling is compared to vanilla SSD, both
|
||||
with and without entropy thresholding.
|
||||
|
||||
@ -655,7 +655,7 @@ on the result.
|
||||
|
||||
Both, vanilla SSD with entropy thresholding and Bayesian SSD with
|
||||
entropy thresholding, were tested for entropy thresholds ranging
|
||||
from 0.1 to 2.5 as specified in Miller et al.~\cite{Miller2018}.
|
||||
from 0.1 to 2.4 inclusive as specified in Miller et al.~\cite{Miller2018}.
|
||||
|
||||
\section{Results}
|
||||
|
||||
|
||||
Reference in New Issue
Block a user