Deep Movement Primitives: Toward Breast Cancer Examination Robot

Breast cancer is the most common type of worldwide. A robotic system performing autonomous breast palpation can make a significant impact on related health sector However, robot programming for palpating with different geometries very complex and unsolved. Robot learning from demonstrations (LfD) reduces time cost. available LfD are lacking modelling manipulation path/trajectory as an explicit function visual sensory information. This paper presents novel approach to planning called deep Movement Primitives that successfully generates movements manipulator reach phantom perform palpation. We show effectiveness our by series real-robot experiments reaching phantom. The experimental results indicate outperforms state-of-the-art method.