Cas - Peal

Enhanced Local Texture Feature Sets for Face Recognition ... - HAL

: Images are captured from multiple angles using a horizontal semicircular camera array to simulate different head rotations. Cas Peal

The database was developed by the Joint Research & Development Laboratory (JDL) of the Chinese Academy of Sciences (CAS). Its primary purpose is to provide researchers with a robust dataset for training and testing algorithms against real-world visual challenges. Core "PEAL" Features Enhanced Local Texture Feature Sets for Face Recognition

is a large-scale Chinese face database created to evaluate face recognition algorithms, specifically focusing on four key types of variations known as PEAL : P ose, E xpression, A ccessories, and L ighting . Its primary purpose is to provide researchers with

Researchers use this data to benchmark various algorithms, including Principal Component Analysis (PCA) and Local Binary Patterns (LBP) , ensuring they can handle the diverse conditions found in practical applications.

: Features images taken under 15 different lighting directions to test an algorithm's robustness to shadows and illumination changes. Database Specifications

: The dataset provides metadata such as eye positions, face pose angles, and expression labels to help refine detection systems.