ID-Sim: An Identity-Focused Similarity Metric

TL;DR

ID-Sim is a new metric designed to better reflect human sensitivity to identity differences in images, aiding progress in identity-focused vision tasks.

Contribution

The paper introduces ID-Sim, a novel identity-focused similarity metric trained on diverse real and synthetic data, with a new benchmark for evaluation.

Findings

ID-Sim aligns closely with human annotations across tasks.

The metric improves identity recognition and retrieval consistency.

It facilitates progress in personalized image generation.

Abstract

Humans have remarkable selective sensitivity to identities -- easily distinguishing between highly similar identities, even across significantly different contexts such as diverse viewpoints or lighting. Vision models have struggled to match this capability, and progress toward identity-focused tasks such as personalized image generation is slowed by a lack of identity-focused evaluation metrics. To help facilitate progress, we propose ID-Sim, a feed-forward metric designed to faithfully reflect human selective sensitivity. To build ID-Sim, we curate a high-quality training set of images spanning diverse real-world domains, augmented with generative synthetic data that provides controlled, fine-grained identity and contextual variations. We evaluate our metric on a new unified evaluation benchmark for assessing consistency with human annotations across identity-focused recognition, retrieval, and generative tasks.