Rezultati pretraživanja
  1. 16. ruj 2019.

    Hybrid BiLSTM/HMM still is the state-of-the-art in acoustic modeling for ASR together with transformer language models, survey talk by Ralf Schlüter

  2. 16. ruj 2019.
  3. 18. ruj 2019.

    Student reception yesterday night

  4. 15. ruj 2019.

    great summary with this table on pros cons of traditional voice convertion methods

  5. 17. ruj 2019.

    Dayan Ribas and Antonio Miguel from vivolab ⁦⁩ are presenting our work on Speech Enhancement with Wide Residual Networks in

  6. 16. ruj 2019.

    Attention! Attention! Attention! It is everywhere

  7. 19. ruj 2019.
  8. 18. ruj 2019.

    We are now presenting our latest works on Speaker Verification and Speech Enhancement at Vivolab ⁦

  9. 15. ruj 2019.

    attending Voice Convertion tutorial

  10. 15. ruj 2019.

    Nice start with a Voice conversion hands-on with WaveNet

  11. 16. ruj 2019.

    A great overview of speech synthesis techniques and evolution is happening now by prof. Tokuda at the main hall

    Prikaži ovu nit
  12. 16. ruj 2019.

    Presenting 2 studies at today 🙂

  13. 15. ruj 2019.
  14. 19. ruj 2019.

    Ready at Hall 10/D to present our poster and demo for our paper “Prosodic phrase alignment for machine dubbing”

  15. ISCA General Assembly / President report John Hansen

  16. 15. ruj 2019.

    Tutorial done! thanks to all the people who attending! very interesting discussions! It was a great pleasure to prepare and present it with ⁦⁩. Time to enjoy Graz (picture taken at the top of the castle) !

  17. 16. ruj 2019.

    PASE, where prior knowledge and deep learning collide to represent speech from waveforms, will be presented at Hall 11 at 12h today!

  18. 15. ruj 2019.

    At ? Check out our poster Tuesday afternoon on the distinctiveness of intonational contours in American English!

    Prikaži ovu nit
  19. are you talking on the phone during lecture? Don't disturb the lecturer by speaking out loud. Tanja Schultz can read your muscles, and convert the signals to speech that's only audible to you and your friend.

  20. 16. ruj 2019.

    Some numbers from Interspeech 2019: * 2056 registered participants (a new record!) * 2180 submissions (>20% more than largest previous Interspeech) * 4509 authors * 1335 reviewers * 49.3% acceptance rate

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.