UFPR-PEs: A Brazilian Face Recognition Benchmark with Self-Declared Race/Color Labels
2026-08-31 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors created UFPR-PEs, a new dataset to test how face recognition systems perform on videos of Brazilian politicians, using official race categories from Brazil's census, including unique groups like 'parda'. Their dataset uses real, compressed public video clips with different quality levels, making it more like what systems face in the real world. They found that how well these systems work depends a lot on the video quality and that differences between race groups should be understood together with how hard the images are to recognize. This work helps researchers study bias in face recognition more fairly and realistically.
face recognitionbias evaluationdatasetrace categoriesBrazilian censusvideo compressionverificationidentificationdemographic reliabilityimage quality
Authors
Alexandre Diano, Bernardo Biesseck, Gabriel Polo, Vinicius Gregorio, Laura Lopes, Diego Addan, David Menotti
Abstract
While face recognition systems are widely deployed, ensuring their demographic reliability and robustness under uncontrolled visual conditions remains a critical challenge. To bridge this gap, we present UFPR-PEs, a benchmark for face recognition bias evaluation using public videos of elected Brazilian politicians annotated with official self-declared race/color categories. The dataset adopts the Brazilian census taxonomy, including the parda category, which has no direct equivalent in the U.S.- or Europe-centric schemas commonly used in prior benchmarks. Our benchmark is built from compressed public video and preserves difficult samples so that performance can be analyzed under realistic conditions. We describe the construction pipeline, report dataset statistics, and evaluate face recognition performance across verification and (closed- and open-set) identification settings, including subgroup analysis by race/color and difficulty level. The results show that recognition performance varies substantially with image quality, and that subgroup gaps must be interpreted jointly with visual difficulty rather than in isolation. Overall, UFPR-PEs provides a reproducible and demographically grounded setting for studying face recognition bias under challenging public video conditions.