Minimizing Residuals for Native-Nonnative Voice Conversion in a Sparse, Anchor-Based Representation of Speech | IEEE Conference Publication | IEEE Xplore