AgingMapGAN (AMGAN): High-Resolution Controllable Face Aging with Spatially-Aware Conditional GANs

TL;DR

AgingMapGAN (AMGAN) introduces a high-resolution face aging method that incorporates ethnicity-specific information and spatial control, improving quality and scalability for real-world applications.

Contribution

The paper presents a novel spatially-aware conditional GAN for high-resolution face aging with ethnicity-specific control and weak spatial supervision.

Findings

Enhanced quality of aged face images.

Effective control over aging process and ethnicity features.

Scalable to high-definition images with limited computational overhead.

Abstract

Existing approaches and datasets for face aging produce results skewed towards the mean, with individual variations and expression wrinkles often invisible or overlooked in favor of global patterns such as the fattening of the face. Moreover, they offer little to no control over the way the faces are aged and can difficultly be scaled to large images, thus preventing their usage in many real-world applications. To address these limitations, we present an approach to change the appearance of a high-resolution image using ethnicity-specific aging information and weak spatial supervision to guide the aging process. We demonstrate the advantage of our proposed method in terms of quality, control, and how it can be used on high-definition images while limiting the computational overhead.