Metaverse Gait Authentication Dataset (MGAD)

The Metaverse Gait Authentication Dataset (MGAD) is a large-scale gait dataset designed for biometric authentication in virtual environments. It contains gait data from 5,000 simulated users, generated in Unity 3D and processed using OpenPose and MediaPipe to extract 16 key features, including stride length, step frequency, joint angles, ground reaction forces, and gait symmetry index. The dataset has undergone extensive preprocessing, feature engineering, and normalization to ensure high-quality, noise-free data suitable for machine learning applications. MGAD is a benchmark for evaluating gait-based authentication models in the Metaverse, enabling researchers to explore biometric security solutions in immersive environments.