Search papers, labs, and topics across Lattice.
This paper introduces ATTN-FIQA, a training-free face image quality assessment (FIQA) method that leverages attention scores from pre-trained Vision Transformer (ViT) based face recognition models. It posits that attention magnitudes inherently encode image quality, with high-quality images producing focused, high-magnitude attention patterns. The method extracts and aggregates pre-softmax attention matrices from the final transformer block to compute image-level quality scores, demonstrating strong correlation with face image quality across eight benchmark datasets and four FR models.
Turns out, your pre-trained face recognition ViT already knows which faces are high quality, just by looking at the attention maps.
Face Image Quality Assessment (FIQA) aims to assess the recognition utility of face samples and is essential for reliable face recognition (FR) systems. Existing approaches require computationally expensive procedures such as multiple forward passes, backpropagation, or additional training, and only recent work has focused on the use of Vision Transformers. Recent studies highlighted that these architectures inherently function as saliency learners with attention patterns naturally encoding spatial importance. This work proposes ATTN-FIQA, a novel training-free approach that investigates whether pre-softmax attention scores from pre-trained Vision Transformer-based face recognition models can serve as quality indicators. We hypothesize that attention magnitudes intrinsically encode quality: high-quality images with discriminative facial features enable strong query-key alignments producing focused, high-magnitude attention patterns, while degraded images generate diffuse, low-magnitude patterns. ATTN-FIQA extracts pre-softmax attention matrices from the final transformer block, aggregate multi-head attention information across all patches, and compute image-level quality scores through simple averaging, requiring only a single forward pass through pre-trained models without architectural modifications, backpropagation, or additional training. Through comprehensive evaluation across eight benchmark datasets and four FR models, this work demonstrates that attention-based quality scores effectively correlate with face image quality and provide spatial interpretability, revealing which facial regions contribute most to quality determination.
