Enhance Convolutional Neural Networks with Noise Incentive Block

TL;DR

This paper introduces the Noise Incentive Block (NIB), a plug-in module that enhances CNNs by mitigating flatness degradation, leading to more vivid and detailed image generation from flat inputs across various tasks.

Contribution

The paper proposes a model-agnostic Noise Incentive Block that breaks flat input conditions in CNNs, improving output quality without altering the original model architecture.

Findings

CNNs with NIB generate more vivid images with richer details.

NIB effectively mitigates flatness degradation in various image generation tasks.

Enhanced models outperform baseline CNNs in visual quality metrics.

Abstract

As a generic modeling tool, Convolutional Neural Networks (CNNs) have been widely employed in image generation and translation tasks. However, when fed with a flat input, current CNN models may fail to generate vivid results due to the spatially shared convolution kernels. We call it the flatness degradation of CNNs. Unfortunately, such degradation is the greatest obstacles to generate a spatially-variant output from a flat input, which has been barely discussed in the previous literature. To tackle this problem, we propose a model agnostic solution, i.e. Noise Incentive Block (NIB), which serves as a generic plug-in for any CNN generation model. The key idea is to break the flat input condition while keeping the intactness of the original information. Specifically, the NIB perturbs the input data symmetrically with a noise map and reassembles them in the feature domain as driven by the objective function. Extensive experiments show that existing CNN models equipped with NIB survive from the flatness degradation and are able to generate visually better results with richer details in some specific image generation tasks given flat inputs, e.g. semantic image synthesis, data-hidden image generation, and deep neural dithering.