Exploring Novel Pooling Strategies for Edge Preserved Feature Maps in Convolutional Neural Networks

TL;DR

This paper introduces two novel edge-preserving pooling strategies for CNNs, demonstrating improved performance over traditional and blur pooling methods in classification, segmentation, and autoencoder tasks.

Contribution

The paper proposes two new pooling methods, LGCA and WADCA, that preserve edges in feature maps, enhancing CNN performance across multiple tasks.

Findings

Proposed pooling methods outperform conventional pooling.

Edge preservation improves segmentation accuracy.

Enhanced autoencoder reconstructions.

Abstract

With the introduction of anti-aliased convolutional neural networks (CNN), there has been some resurgence in relooking the way pooling is done in CNNs. The fundamental building block of the anti-aliased CNN has been the application of Gaussian smoothing before the pooling operation to reduce the distortion due to aliasing thereby making CNNs shift invariant. Wavelet based approaches have also been proposed as a possibility of additional noise removal capability and gave interesting results for even segmentation tasks. However, all the approaches proposed completely remove the high frequency components under the assumption that they are noise. However, by removing high frequency components, the edges in the feature maps are also smoothed. In this work, an exhaustive analysis of the edge preserving pooling options for classification, segmentation and autoencoders are presented. Two novel pooling approaches are presented such as Laplacian-Gaussian Concatenation with Attention (LGCA) pooling and Wavelet based approximate-detailed coefficient concatenation with attention (WADCA) pooling. The results suggest that the proposed pooling approaches outperform the conventional pooling as well as blur pooling for classification, segmentation and autoencoders.