RoutingGAN: Routing Age Progression and Regression with Disentangled Learning

TL;DR

RoutingGAN introduces a novel routing mechanism in GANs to effectively disentangle and learn age progression and regression effects within a single model, improving flexibility and performance.

Contribution

The paper proposes RoutingGAN, a new method that routes effects in high-level features, enabling simultaneous learning of multiple age effects with shared and specific filters in one model.

Findings

Outperforms existing methods qualitatively and quantitatively

Effectively disentangles age-invariant features from input faces

Learns various age effects within a single model

Abstract

Although impressive results have been achieved for age progression and regression, there remain two major issues in generative adversarial networks (GANs)-based methods: 1) conditional GANs (cGANs)-based methods can learn various effects between any two age groups in a single model, but are insufficient to characterize some specific patterns due to completely shared convolutions filters; and 2) GANs-based methods can, by utilizing several models to learn effects independently, learn some specific patterns, however, they are cumbersome and require age label in advance. To address these deficiencies and have the best of both worlds, this paper introduces a dropout-like method based on GAN~(RoutingGAN) to route different effects in a high-level semantic feature space. Specifically, we first disentangle the age-invariant features from the input face, and then gradually add the effects to the features by residual routers that assign the convolution filters to different age groups by dropping out the outputs of others. As a result, the proposed RoutingGAN can simultaneously learn various effects in a single model, with convolution filters being shared in part to learn some specific effects. Experimental results on two benchmarked datasets demonstrate superior performance over existing methods both qualitatively and quantitatively.