1: \begin{abstract}

2: Normalization layers have been shown to improve convergence in deep neural networks, and even add useful inductive biases.  %Normalizing along the batch dimension as done by \gls{bn} introduces issues caused by inaccurate batch statistics estimation when the batch size is small. \gls{gn} was recently proposed to address this issue by normalizing based on predefined groups of filters.

3: In many vision applications the local spatial context of the features is important, but most common normalization schemes including \gls{gn}, \gls{in}, and \gls{ln} normalize over the entire spatial dimension of a feature. This can wash out important signals and degrade performance. For example, in applications that use satellite imagery, input images can be arbitrarily large; consequently, it is nonsensical to normalize over the entire area. \gls{pn}, on the other hand, only normalizes over a single spatial position at a time. A natural compromise is to normalize features by local context, while also taking into account group level information. In this paper, we propose {\em \gls{lcn}}: a normalization layer where every feature is normalized based on a window around it and the filters in its group. %\todo{I found that local response normalization is commonly interchanged with local contrast norm which uses (LCN), should we change this}

4: We propose an algorithmic solution to make \gls{lcn} efficient for arbitrary window sizes, even if every point in the image has a unique window. \gls{lcn} outperforms its \gls{bn}, \gls{gn}, \gls{in}, and \gls{ln} counterparts for object detection, semantic segmentation, and instance segmentation applications in several benchmark datasets, while keeping performance independent of the batch size and facilitating transfer learning.

5: \end{abstract}

6: