Markerless (ML) gait analysis with multi or single camera approaches is increasingly investigated as an accessible alternative to marker-based (MB) motion capture, as it avoids the placement of physical markers on the subject’s skin and simplifies data acquisition. Current dual camera solutions include OpenCap, an open-source application from Stanford University, which reconstructs 3D body keypoints from two iPhones, predicts 3D virtual markers through marker augmentation (AGM) and estimates joint kinematics through OpenSim inverse kinematics (IK). Conversely, single-camera approaches based on statistical body-model fitting provide 3D joint centres from RGB/RGB-D data, but not the anatomical landmarks required for the Plug-in Gait (PiG) biomechanical protocol to estimate 3D joint kinematics.