 This paper presents a novel human-machine interaction, HMI, recognition wearable system for continuous multiple degree of freedom, DOF, ankle-foot movements. The system uses sole kinematic signals from inertial measurement units and multi-class support vector machines, SVMs, to achieve high-level control of multiple DOF wearable lower limb robotics. The system was validated through experiments with multiple participants achieving overall accuracies of 98.45% and 85.3% respectively within a prediction time of 10.9 ms and 14.1 ms. These results demonstrate that the proposed system is more effective at recognizing multiple continuous DOF foot movements than existing strategies and can fill the HMI high-level control demand for multiple DOF wearable lower limb robotics. This article was authored by Chen Xiaoshu, Lan Luo, Jingying Mai, and others.