Parallelized and pipelined hardware implementation of computationally expensive prediction filters | Publicación