FROQ: Observing Face Recognition Models for Efficient Quality Assessment

2025-09-22

AI Summary

Face Image Quality Assessment (FIQA) faces a dual bottleneck: supervised methods require large-scale human annotations, while unsupervised approaches suffer from low efficiency and insufficient accuracy. To address this, we propose FROQ—a training-free, semi-supervised FIQA method that uniquely integrates unsupervised pseudo-label generation with intermediate-layer feature analysis of pre-trained face recognition models. Specifically, FROQ generates pseudo-quality labels via input perturbations and leverages internal model representations to calibrate and uncover discriminative quality cues. Importantly, FROQ is model-agnostic, requiring no adaptation or fine-tuning of any modern face recognition architecture. Extensive experiments across four state-of-the-art face recognition backbones and eight benchmark datasets demonstrate that FROQ consistently outperforms existing unsupervised and weakly supervised FIQA methods, achieving new state-of-the-art performance in both accuracy and inference speed.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Qualitative Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Face Recognition (FR) plays a crucial role in many critical (high-stakes) applications, where errors in the recognition process can lead to serious consequences. Face Image Quality Assessment (FIQA) techniques enhance FR systems by providing quality estimates of face samples, enabling the systems to discard samples that are unsuitable for reliable recognition or lead to low-confidence recognition decisions. Most state-of-the-art FIQA techniques rely on extensive supervised training to achieve accurate quality estimation. In contrast, unsupervised techniques eliminate the need for additional training but tend to be slower and typically exhibit lower performance. In this paper, we introduce FROQ (Face Recognition Observer of Quality), a semi-supervised, training-free approach that leverages specific intermediate representations within a given FR model to estimate face-image quality, and combines the efficiency of supervised FIQA models with the training-free approach of unsupervised methods. A simple calibration step based on pseudo-quality labels allows FROQ to uncover specific representations, useful for quality assessment, in any modern FR model. To generate these pseudo-labels, we propose a novel unsupervised FIQA technique based on sample perturbations. Comprehensive experiments with four state-of-the-art FR models and eight benchmark datasets show that FROQ leads to highly competitive results compared to the state-of-the-art, achieving both strong performance and efficient runtime, without requiring explicit training.

Problem

• Develops training-free face quality assessment using FR model representations
• Combines supervised method efficiency with unsupervised training-free approach
• Generates pseudo-labels through novel perturbation-based FIQA technique

Innovation

• Leverages intermediate representations for quality assessment
• Uses pseudo-quality labels from sample perturbations
• Combines supervised efficiency with training-free approach