ResearchBenchmarksSpeech Synthesis on Internal US English female (evaluation)For AI assistantsMarkdownContent indexSpeech Synthesis on Internal US English female (evaluation)Follow0.048CMOSOur Model-0.001920.011040.0240.03696Sep 19, 2018Evaluation resultsMethodMethodLinksCMOSMOSOur ModelInput=phonemes, Vocode...Input=phonemes, Vocoder=WaveNet2018.090.048—Tacotron2Input=phonemes, Vocode...Input=phonemes, Vocoder=WaveNet2018.090—