Truncated error correction for flexible approximate multiplication | IEEE Conference Publication | IEEE Xplore