AI People Wiki

Builders & open source

RJ Skerry-Ryan

Researcher listed in Google Research's people directory (publications 2017–2025), Google

  • Researcher listed in Google Research's people directory (publications 2017–2025), Google Research1

RJ Skerry-Ryan is listed in Google Research's people directory, with Google publications from 2017 through 2025 and speech processing among the research areas. Skerry-Ryan is the second author of Tacotron: Towards End-to-End Speech Synthesis, an end-to-end generative text-to-speech model whose paper lists its authors at Google, and a co-author of the Tacotron 2 paper. Skerry-Ryan is also first author of Towards End-to-End Prosody Transfer for Expressive Speech Synthesis with Tacotron, which extends Tacotron with a learned latent embedding space of prosody.12345

At a glance

Current role
Researcher listed in Google Research's people directory (publications 2017–2025)
Company
Google
Sector
Company
Work listed
3
Sources cited
8
Profile reviewed

Research

Skerry-Ryan's research at Google centers on speech synthesis, and the Google Research people page lists machine intelligence and machine perception among Skerry-Ryan's research areas. Skerry-Ryan was second author of the 2017 Tacotron paper, an end-to-end generative text-to-speech model, and a co-author of Tacotron 2 later that year.1234

In March 2018 Skerry-Ryan was first author of a paper extending Tacotron with a learned latent embedding space of prosody for expressive speech, and co-authored the Style Tokens paper on unsupervised style modeling, control and transfer. Later work on Tacotron includes location-relative attention mechanisms for robust long-form synthesis (2019) and Wave-Tacotron (2020), which generates speech waveforms directly from text without spectrograms.5678

More recent publications on Skerry-Ryan's Google Research page include speaker generation with TacoSpawn (ICASSP 2022), Spectron, which adapts pretrained language models for spoken question answering and speech continuation (ICLR 2024), and ZeroBAS, a method for synthesizing binaural speech from monaural recordings (Interspeech 2025).1

Work and contributions

  • Tacotron Published research

    Second author of Tacotron: Towards End-to-End Speech Synthesis (2017), which presents a text-to-speech model that synthesizes speech directly from characters; the paper lists the authors at Google, Inc.23

  • Tacotron prosody transfer Published research

    First author of Towards End-to-End Prosody Transfer for Expressive Speech Synthesis with Tacotron (2018), which conditions Tacotron on a learned prosody embedding derived from a reference recording.5

  • Tacotron 2 Published research

    Co-author of Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions (2017), the Tacotron 2 paper.4

Timeline

  1. Posted the Tacotron end-to-end speech synthesis paper as second author.2
  2. Co-authored the Tacotron 2 paper.4
  3. Co-authored "Style Tokens" on unsupervised style modeling in end-to-end speech synthesis.6
  4. Posted the Tacotron prosody transfer paper as first author.5
  5. Co-authored a paper on location-relative attention for robust long-form speech synthesis.7
  6. Co-authored Wave-Tacotron, a spectrogram-free end-to-end text-to-speech model.8
  7. Co-authored "Speaker Generation", introducing TacoSpawn, at ICASSP 2022.1
  8. Co-authored Spectron, on spoken question answering with a spectrogram-powered language model, at ICLR 2024.1
  9. Co-authored "Zero-Shot Mono-to-Binaural Speech Synthesis" at Interspeech 2025.1

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

8 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. RJ Skerry-Ryan — Google Research people pageGoogle Research · Official page · accessed Oct 21, 2026
  2. Tacotron: Towards End-to-End Speech Synthesis (arXiv 1703.10135)arXiv · Paper · published Mar 29, 2017 · accessed Oct 21, 2026
  3. Tacotron: Towards End-to-End Speech Synthesis — paper PDF (author affiliation: Google, Inc.)arXiv · Paper · published Mar 29, 2017 · accessed Oct 21, 2026
  4. Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions (arXiv 1712.05884)arXiv · Paper · published Dec 16, 2017 · accessed Oct 21, 2026
  5. Towards End-to-End Prosody Transfer for Expressive Speech Synthesis with Tacotron (arXiv 1803.09047)arXiv · Paper · published Mar 24, 2018 · accessed Oct 21, 2026
  6. Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis (arXiv 1803.09017)arXiv · Paper · published Mar 23, 2018 · accessed Oct 21, 2026
  7. Location-Relative Attention Mechanisms For Robust Long-Form Speech Synthesis (arXiv 1910.10288)arXiv · Paper · published Oct 23, 2019 · accessed Oct 21, 2026
  8. Wave-Tacotron: Spectrogram-free end-to-end text-to-speech synthesis (arXiv 2011.03568)arXiv · Paper · published Nov 6, 2020 · accessed Oct 21, 2026