Coloring Deep CNN Layers with Activation Hue Loss

TL;DR

This paper introduces the activation hue, a novel angular regularization method for deep CNNs, which improves classification performance by encouraging class-specific activation angles during training.

Contribution

It proposes the activation hue concept and a regularization loss based on angular activation labels, enhancing CNN training effectiveness.

Findings

Activation hue captures class-informative activation angles.

Regularization with hue loss improves classification accuracy.

Method is effective across various datasets including ImageNet.

Abstract

This paper proposes a novel hue-like angular parameter to model the structure of deep convolutional neural network (CNN) activation space, referred to as the {\em activation hue}, for the purpose of regularizing models for more effective learning. The activation hue generalizes the notion of color hue angle in standard 3-channel RGB intensity space to $N$-channel activation space. A series of observations based on nearest neighbor indexing of activation vectors with pre-trained networks indicate that class-informative activations are concentrated about an angle $\theta$ in both the $(x,y)$ image plane and in multi-channel activation space. A regularization term in the form of hue-like angular $\theta$ labels is proposed to complement standard one-hot loss. Training from scratch using combined one-hot + activation hue loss improves classification performance modestly for a wide variety of classification tasks, including ImageNet.