A Lightweight Ensemble-Based Face Image Quality Assessment Method with Correlation-Aware Loss

2025-09-12

AI Summary

Existing no-reference facial image quality assessment (FIQA) methods suffer from insufficient modeling of face-specific degradations and high computational overhead. To address these issues, this paper proposes a lightweight and efficient FIQA framework. Methodologically, it integrates MobileNetV3-Small and ShuffleNetV2 into a compact ensemble architecture, employs average fusion at the prediction layer, and introduces a correlation-aware loss function—MSECorrLoss—that jointly optimizes mean squared error and Pearson correlation coefficient—thereby significantly improving alignment between predicted scores and human subjective judgments. Evaluated on the VQualA benchmark, the model achieves state-of-the-art performance with Spearman and Pearson correlation coefficients of 0.9829 and 0.9894, respectively, while reducing model parameters and FLOPs substantially. This enables practical deployment in resource-constrained scenarios.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Qualitative Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Face image quality assessment (FIQA) plays a critical role in face recognition and verification systems, especially in uncontrolled, real-world environments. Although several methods have been proposed, general-purpose no-reference image quality assessment techniques often fail to capture face-specific degradations. Meanwhile, state-of-the-art FIQA models tend to be computationally intensive, limiting their practical applicability. We propose a lightweight and efficient method for FIQA, designed for the perceptual evaluation of face images in the wild. Our approach integrates an ensemble of two compact convolutional neural networks, MobileNetV3-Small and ShuffleNetV2, with prediction-level fusion via simple averaging. To enhance alignment with human perceptual judgments, we employ a correlation-aware loss (MSECorrLoss), combining mean squared error (MSE) with a Pearson correlation regularizer. Our method achieves a strong balance between accuracy and computational cost, making it suitable for real-world deployment. Experiments on the VQualA FIQA benchmark demonstrate that our model achieves a Spearman rank correlation coefficient (SRCC) of 0.9829 and a Pearson linear correlation coefficient (PLCC) of 0.9894, remaining within competition efficiency constraints.

Problem

• Lightweight face image quality assessment method
• Addresses face-specific degradations in uncontrolled environments
• Balances accuracy with computational efficiency

Innovation

• Lightweight ensemble of compact CNNs
• Prediction-level fusion via averaging
• Correlation-aware loss combining MSE and Pearson