Updated method part with current Bayesian filtering process

Signed-off-by: Jim Martens <github@2martens.de>
This commit is contained in:
Jim Martens 2019-09-09 12:08:24 +02:00
parent bd699f21c3
commit 9fde66abf1
1 changed files with 3 additions and 1 deletions

View File

@ -574,7 +574,9 @@ varying classifications are averaged into multiple lower confidence
values which should increase the entropy and, hence, flag an
observation for removal.
The final step is a top \(k\) selection.
The remainder of the filtering follows the vanilla SSD procedure: per-class
confidence threshold, non-maximum suppression, and a top \(k\) selection
at the end.
\chapter{Experimental Setup and Results}