I want to know whether openCV version 3.0 provides parallel GPU implementation of face detection i.e with CUDA ? If it provides that then how the training of haar cascade is done for it ? Please direct me to reading , online materials which can help me better understand the concept behind this.