Pro-Pose: Unpaired Full-Body Portrait Synthesis via Canonical UV Maps

TL;DR

This paper introduces a novel method to convert in-the-wild photographs into standardized, high-fidelity avatars by transforming images into a canonical UV space, enabling pose manipulation and identity preservation without paired datasets.

Contribution

We propose a UV space transformation and reposing technique that decouples pose from appearance, allowing unpaired data to be used for high-quality avatar synthesis and identity retention.

Findings

Achieves high-quality reposed portraits with strong identity preservation.

Improves downstream applications like Virtual Try-On significantly.

Operates effectively without paired datasets by leveraging UV space.

Abstract

Photographs of people taken by professional photographers typically present the person in beautiful lighting, with an interesting pose, and flattering quality. This is unlike common photos people take of themselves in uncontrolled conditions. In this paper, we explore how to canonicalize a person's 'in-the-wild' photograph into a controllable, high-fidelity avatar -- reposed in a simple environment with standardized minimal clothing. A key challenge is preserving the person's unique whole-body identity, facial features, and body shape while stripping away the complex occlusions of their original garments. While a large paired dataset of the same person in varied clothing and poses would simplify this, such data does not exist. To that end, we propose two key insights: 1) Our method transforms the input photo into a canonical full-body UV space, which we couple with a novel reposing methodology to model occlusions and synthesize novel views. Operating in UV space allows us to decouple pose from appearance and leverage massive unpaired datasets. 2) We personalize the output photo via multi-image finetuning to ensure robust identity preservation under extreme pose changes. Our approach yields high-quality, reposed portraits that achieve strong quantitative performance on real-world imagery, providing an ideal, clean biometric canvas that significantly improves the fidelity of downstream applications like Virtual Try-On (VTO).