Comparison of two methods for computing flat outputs | IEEE Conference Publication | IEEE Xplore