From 63663df44871970b2b0d30fee958fbea86009d09 Mon Sep 17 00:00:00 2001 From: Simon Meister Date: Mon, 13 Nov 2017 13:59:43 +0100 Subject: [PATCH] WIP --- approach.tex | 33 +++++++++++++++++++-------------- background.tex | 2 +- 2 files changed, 20 insertions(+), 15 deletions(-) diff --git a/approach.tex b/approach.tex index 0a43dcd..5d59b63 100644 --- a/approach.tex +++ b/approach.tex @@ -39,13 +39,13 @@ $o_t^{cam}$& softmax, 2 & 1 $\times$ 2 \\ \midrule \multicolumn{3}{c}{\textbf{RoI Head: Motions}}\\ \midrule -& From M$_0$: flatten & N$_{RPN}$ $\times$ 7 $\cdot$ 7 $\cdot$ 256 \\ -T$_1$ & $\begin{bmatrix}\textrm{fully connected}, 1024\end{bmatrix}$ $\times$ 2 & N$_{RPN}$ $\times$ 1024 \\ -$\forall k: R_t^k$ & From T$_1$: fully connected, 3 & N$_{RPN}$ $\times$ 3 \\ -$\forall k: t_t^k$ & From T$_1$: fully connected, 3 & N$_{RPN}$ $\times$ 3 \\ -$\forall k: p_t^k$ & From T$_1$: fully connected, 3 & N$_{RPN}$ $\times$ 3 \\ -& From T$_1$: fully connected, 2 & N$_{RPN}$ $\times$ 2 \\ -$\forall k: o_t^k$ & softmax, 2 & N$_{RPN}$ $\times$ 2 \\ +& From M$_0$: flatten & N$_{RoI}$ $\times$ 7 $\cdot$ 7 $\cdot$ 256 \\ +T$_1$ & $\begin{bmatrix}\textrm{fully connected}, 1024\end{bmatrix}$ $\times$ 2 & N$_{RoI}$ $\times$ 1024 \\ +$\forall k: R_t^k$ & From T$_1$: fully connected, 3 & N$_{RoI}$ $\times$ 3 \\ +$\forall k: t_t^k$ & From T$_1$: fully connected, 3 & N$_{RoI}$ $\times$ 3 \\ +$\forall k: p_t^k$ & From T$_1$: fully connected, 3 & N$_{RoI}$ $\times$ 3 \\ +& From T$_1$: fully connected, 2 & N$_{RoI}$ $\times$ 2 \\ +$\forall k: o_t^k$ & softmax, 2 & N$_{RoI}$ $\times$ 2 \\ \bottomrule \end{tabular} @@ -88,13 +88,13 @@ $o_t^{cam}$& softmax, 2 & 1 $\times$ 2 \\ \midrule \multicolumn{3}{c}{\textbf{RoI Head: Motions}}\\ \midrule -& From M$_1$: flatten & N$_{RPN}$ $\times$ 14 $\cdot$ 14 $\cdot$ 256 \\ -T$_3$ & $\begin{bmatrix}\textrm{fully connected}, 1024\end{bmatrix}$ $\times$ 2 & N$_{RPN}$ $\times$ 1024 \\ -$\forall k: R_t^k$ & From T$_3$: fully connected, 3 & N$_{RPN}$ $\times$ 3 \\ -$\forall k: t_t^k$ & From T$_3$: fully connected, 3 & N$_{RPN}$ $\times$ 3 \\ -$\forall k: p_t^k$ & From T$_3$: fully connected, 3 & N$_{RPN}$ $\times$ 3 \\ -& From T$_2$: fully connected, 2 & N$_{RPN}$ $\times$ 2 \\ -$\forall k: o_t^k$ & softmax, 2 & N$_{RPN}$ $\times$ 2 \\ +& From M$_1$: flatten & N$_{RoI}$ $\times$ 14 $\cdot$ 14 $\cdot$ 256 \\ +T$_3$ & $\begin{bmatrix}\textrm{fully connected}, 1024\end{bmatrix}$ $\times$ 2 & N$_{RoI}$ $\times$ 1024 \\ +$\forall k: R_t^k$ & From T$_3$: fully connected, 3 & N$_{RoI}$ $\times$ 3 \\ +$\forall k: t_t^k$ & From T$_3$: fully connected, 3 & N$_{RoI}$ $\times$ 3 \\ +$\forall k: p_t^k$ & From T$_3$: fully connected, 3 & N$_{RoI}$ $\times$ 3 \\ +& From T$_2$: fully connected, 2 & N$_{RoI}$ $\times$ 2 \\ +$\forall k: o_t^k$ & softmax, 2 & N$_{RoI}$ $\times$ 2 \\ \bottomrule \end{tabular} @@ -324,6 +324,11 @@ loss could benefit motion regression by removing any loss balancing issues betwe rotation, translation and pivot terms \cite{PoseNet2}, which can make it interesting even when 3D motion ground truth is available. +\subsection{Inference} +\label{ssec:inference} +During inference, we proceed analogously to Mask R-CNN. +In the same way as the RoI mask head, at test time, we compute the RoI motion head +from the features extracted with refined bounding boxes. \subsection{Dense flow from motion} \label{ssec:postprocessing} diff --git a/background.tex b/background.tex index 0808417..db8dcca 100644 --- a/background.tex +++ b/background.tex @@ -439,7 +439,7 @@ Figure from \cite{FPN}. \label{figure:fpn_block} \end{figure} -\subsection{Training Mask R-CNN} +\subsection{Mask R-CNN: Training and Inference} \paragraph{Loss definitions} For regression, we define the smooth $\ell_1$ regression loss as \begin{equation}