From 96f9a696f5ab2121b0f464d1937e92180fe27507 Mon Sep 17 00:00:00 2001 From: Jim Martens Date: Tue, 1 Oct 2019 13:30:30 +0200 Subject: [PATCH] Made background section about dropout sampling more generalised Signed-off-by: Jim Martens --- body.tex | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/body.tex b/body.tex index 53c11a7..4615052 100644 --- a/body.tex +++ b/body.tex @@ -384,7 +384,7 @@ the \gls{entropy} \(H(\mathbf{q}) = - \sum_i q_i \cdot \log q_i\). Miller et al.~\cite{Miller2018} apply the dropout sampling to object detection. In that case \(\mathbf{W}\) represents the -learned weights of a detection network like \gls{SSD}~\cite{Liu2016}. +learned weights of a detection network, for example \gls{SSD}~\cite{Liu2016}. Every forward pass uses a different network \(\widetilde{\mathbf{W}}\) which is approximately sampled from \(p(\mathbf{W}|\mathbf{T})\). Each forward pass in object