Infant fingerprint recognition is unreliable in practice because babies have finer ridge patterns and cannot hold still during capture. Existing quality metrics, designed for adult fingerprints, assume feature distributions that do not hold for infants. A new preprint from researchers including Anil K. Jain evaluates ten fingerprint quality assessment methods on a longitudinal dataset spanning six age groups from newborns to three months, and introduces a parameter-tuning routine to recalibrate methods for age-specific characteristics.

The authors compared default and tuned configurations through recognition experiments, focusing on quality-guided gallery selection. They report that tuned parameter sets outperform default configurations for the majority of methods and age groups, indicating that targeted recalibration improves gallery utility. They also examined template-fusion strategies and latent fingerprint recognition technologies to estimate the best matching rates achievable by the top-performing method.

Because the paper is a single preprint, its findings have not yet been peer-reviewed. The results suggest that existing fingerprint frameworks can be adapted rather than replaced, but the practical gains will depend on replication in larger and more diverse infant populations.