StyleID: A Perception-Aware Dataset and Metric for Stylization-Agnostic Facial Identity Recognition

TL;DR

This paper introduces StyleID, a perception-aware dataset and evaluation framework for facial identity recognition across various stylized portraits, addressing the brittleness of existing encoders under stylization.

Contribution

It provides a new dataset and evaluation method that align face recognition models with human perception across different artistic styles and strengths.

Findings

Fine-tuning encoders with StyleID improves correlation with human judgments.

Calibrated models show increased robustness to out-of-domain, artist-drawn portraits.

StyleID datasets and models are publicly available for research use.

Abstract

Creative face stylization aims to render portraits in diverse visual idioms such as cartoons, sketches, and paintings while retaining recognizable identity. However, current identity encoders, which are typically trained and calibrated on natural photographs, exhibit severe brittleness under stylization. They often mistake changes in texture or color palette for identity drift or fail to detect geometric exaggerations. This reveals the lack of a style-agnostic framework to evaluate and supervise identity consistency across varying styles and strengths. To address this gap, we introduce StyleID, a human perception-aware dataset and evaluation framework for facial identity under stylization. StyleID comprises two datasets: (i) StyleBench-H, a benchmark that captures human same-different verification judgments across diffusion- and flow-matching-based stylization at multiple style strengths, and (ii) StyleBench-S, a supervision set derived from psychometric recognition-strength curves obtained through controlled two-alternative forced-choice (2AFC) experiments. Leveraging StyleBench-S, we fine-tune existing semantic encoders to align their similarity orderings with human perception across styles and strengths. Experiments demonstrate that our calibrated models yield significantly higher correlation with human judgments and enhanced robustness for out-of-domain, artist drawn portraits. All of our datasets, code, and pretrained models are publicly available at https://kwanyun.github.io/StyleID_page/