Recursive Chaining of Reversible Image-to-image Translators For Face Aging

TL;DR

This paper introduces a recursive chaining method of reversible image-to-image translators to model face aging over multiple stages without requiring individual-specific aging data, improving flexibility and interpretability.

Contribution

It proposes a novel recursive chaining of reversible transformers for face aging, enabling multi-phase translation with no individual aging samples and revealing underlying physical aging processes.

Findings

Achieves state-of-the-art face aging results visually.

Allows recursive application of transformers for multi-phase aging.

Uncovers insights into physical aging processes.

Abstract

This paper addresses the modeling and simulation of progressive changes over time, such as human face aging. By treating the age phases as a sequence of image domains, we construct a chain of transformers that map images from one age domain to the next. Leveraging recent adversarial image translation methods, our approach requires no training samples of the same individual at different ages. Here, the model must be flexible enough to translate a child face to a young adult, and all the way through the adulthood to old age. We find that some transformers in the chain can be recursively applied on their own output to cover multiple phases, compressing the chain. The structure of the chain also unearths information about the underlying physical process. We demonstrate the performance of our method with precise and intuitive metrics, and visually match with the face aging state-of-the-art.