Context Normalization Layer with Applications
This work addresses a specific problem in deep learning normalization for image processing, offering an incremental improvement over existing techniques.
The paper tackles the limitations of batch normalization in deep neural networks, such as dependency on mini-batch size and ignoring data clusters, by proposing context normalization for image data, which adapts feature scaling per sample to improve convergence speed and performance, as demonstrated on various datasets.
Normalization is a pre-processing step that converts the data into a more usable representation. As part of the deep neural networks (DNNs), the batch normalization (BN) technique uses normalization to address the problem of internal covariate shift. It can be packaged as general modules, which have been extensively integrated into various DNNs, to stabilize and accelerate training, presumably leading to improved generalization. However, the effect of BN is dependent on the mini-batch size and it does not take into account any groups or clusters that may exist in the dataset when estimating population statistics. This study proposes a new normalization technique, called context normalization, for image data. This approach adjusts the scaling of features based on the characteristics of each sample, which improves the model's convergence speed and performance by adapting the data values to the context of the target task. The effectiveness of context normalization is demonstrated on various datasets, and its performance is compared to other standard normalization techniques.