Meet-in-the-middle: Multi-scale upsampling and matching for cross-resolution face recognition
This addresses a critical problem in surveillance and security by enabling reliable face identification from low-quality sources, though it is incremental as it builds on existing techniques.
The paper tackled cross-resolution face recognition between high-quality portraits and low-quality surveillance images by proposing a method combining super-resolution, resolution matching, and multi-scale template accumulation, achieving state-of-the-art performance on the SCFace dataset without fine-tuning.
In this paper, we aim to address the large domain gap between high-resolution face images, e.g., from professional portrait photography, and low-quality surveillance images, e.g., from security cameras. Establishing an identity match between disparate sources like this is a classical surveillance face identification scenario, which continues to be a challenging problem for modern face recognition techniques. To that end, we propose a method that combines face super-resolution, resolution matching, and multi-scale template accumulation to reliably recognize faces from long-range surveillance footage, including from low quality sources. The proposed approach does not require training or fine-tuning on the target dataset of real surveillance images. Extensive experiments show that our proposed method is able to outperform even existing methods fine-tuned to the SCFace dataset.