outline suzero mml talk
play

Outline: SUZero MML Talk Interspeech talk (for Ewald) Explain one - PowerPoint PPT Presentation

Outline: SUZero MML Talk Interspeech talk (for Ewald) Explain one technique in a bit more detail The experience of a coding sprint Unsupervised acoustic unit discovery for speech synthesis using discrete latent-variable neural


  1. Planned structure • Original idea: Arnu had a sprint for some other work • Duration: Two weeks (probably longer, but then you can leave) • Two teams, with tech support (Elan) and someone crying (Herman) Waveform • Compression team: Symbol-to-speech module FFTNet Vocoder Arnu , Ryan, Andr´ e, Leanne y 1: T ˆ Filterbanks Decoder • Synthesis team: Compression model Ewald , Benji, Lisa, Avashna z 1: N Embed Speaker ID Discretise • Teams would work in parallel h 1: N Encoder x 1: T MFCCs 22 / 35

  2. Planned structure • Herman talks to team leaders every day • Daily stand-ups within each of the teams • Slack: All communication • Trello: Track tasks with boards (backlog, in-progress, done) • Bitbucket: Version control using git, pull requests need to be reviewed 23 / 35

  3. Promises beforehand: What will you get from this? 24 / 35

  4. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) 24 / 35

  5. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun 24 / 35

  6. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! 24 / 35

  7. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills 24 / 35

  8. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe 24 / 35

  9. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe • Do something in a group 24 / 35

  10. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe • Do something in a group • Learn where the DSP and MediaLabs are 24 / 35

  11. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe • Do something in a group • Learn where the DSP and MediaLabs are • Maybe a paper 24 / 35

  12. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe • Do something in a group • Learn where the DSP and MediaLabs are • Maybe a paper . . . probably 24 / 35

  13. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe • Do something in a group • Learn where the DSP and MediaLabs are • Maybe a paper . . . probably . . . almost certainly 24 / 35

  14. Promises beforehand: What will you get from this? • You can leave after two weeks (check with your supervisors) • Have fun • Learn something about speech! • Learn some software engineering skills . . . maybe • Do something in a group • Learn where the DSP and MediaLabs are • Maybe a paper . . . probably . . . almost certainly • Worst case: Pizza and beer 24 / 35

  15. What actually happened 27 / 35

  16. What actually happened • Two teams 27 / 35

  17. What actually happened • Two teams ∼ 27 / 35

  18. What actually happened • Two teams ∼ • Herman talks to team leaders every day � 27 / 35

  19. What actually happened • Two teams ∼ • Herman talks to team leaders every day � • Daily stand-ups within each of the teams � 27 / 35

  20. What actually happened • Two teams ∼ • Herman talks to team leaders every day � • Daily stand-ups within each of the teams � • Slack: All communication 27 / 35

  21. What actually happened • Two teams ∼ • Herman talks to team leaders every day � • Daily stand-ups within each of the teams � • Slack: All communication :( 27 / 35

  22. What actually happened • Two teams ∼ • Herman talks to team leaders every day � • Daily stand-ups within each of the teams � • Slack: All communication :( • Some people didn’t respond; different time zones complicated things 27 / 35

  23. 28 / 35

  24. 29 / 35

  25. 30 / 35

  26. What actually happened • You can leave after two weeks � 31 / 35

  27. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . 31 / 35

  28. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline 31 / 35

  29. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( 31 / 35

  30. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun 31 / 35

  31. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ 31 / 35

  32. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ • Learn something about speech! � 31 / 35

  33. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ • Learn something about speech! � • Learn some software engineering skills . . . maybe � 31 / 35

  34. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ • Learn something about speech! � • Learn some software engineering skills . . . maybe � • Do something in a group � 31 / 35

  35. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ • Learn something about speech! � • Learn some software engineering skills . . . maybe � • Do something in a group � • Learn where the DSP and MediaLabs are :( 31 / 35

  36. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ • Learn something about speech! � • Learn some software engineering skills . . . maybe � • Do something in a group � • Learn where the DSP and MediaLabs are :( • Maybe a paper . . . probably . . . almost certainly � 31 / 35

  37. What actually happened • You can leave after two weeks � • But for Ryan, Andr´ e, Benji . . . almost two months, up to day of submission deadline :( • Have fun ∼ • Learn something about speech! � • Learn some software engineering skills . . . maybe � • Do something in a group � • Learn where the DSP and MediaLabs are :( • Maybe a paper . . . probably . . . almost certainly � • Pizza and beer � 31 / 35

  38. What we learned: Things we did that worked 32 / 35

  39. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) 32 / 35

  40. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) • Role assignment beforehand 32 / 35

  41. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) • Role assignment beforehand • Make expectations clear upfront (e.g. authors on paper and order) 32 / 35

  42. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) • Role assignment beforehand • Make expectations clear upfront (e.g. authors on paper and order) • Using team leaders to deal with big team 32 / 35

  43. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) • Role assignment beforehand • Make expectations clear upfront (e.g. authors on paper and order) • Using team leaders to deal with big team • Flexible in restructuring things on the fly (based on listening to recommendations from team) 32 / 35

  44. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) • Role assignment beforehand • Make expectations clear upfront (e.g. authors on paper and order) • Using team leaders to deal with big team • Flexible in restructuring things on the fly (based on listening to recommendations from team) • Pizza and beer 32 / 35

  45. What we learned: Things we did that worked • Planning beforehand (Herman did some prototyping) • Role assignment beforehand • Make expectations clear upfront (e.g. authors on paper and order) • Using team leaders to deal with big team • Flexible in restructuring things on the fly (based on listening to recommendations from team) • Pizza and beer (Gino’s delivers) 32 / 35

  46. What we learned: Things we did that didn’t work 33 / 35

  47. What we learned: Things we did that didn’t work • Some roles weren’t clear enough (especially for first year masters students) 33 / 35

  48. What we learned: Things we did that didn’t work • Some roles weren’t clear enough (especially for first year masters students) • Some people had other stuff going on in the first two weeks 33 / 35

  49. What we learned: Things we did that didn’t work • Some roles weren’t clear enough (especially for first year masters students) • Some people had other stuff going on in the first two weeks • We focussed on intermediate evaluations which turned out not be that important in the end 33 / 35

  50. What we learned: Things we did that didn’t work • Some roles weren’t clear enough (especially for first year masters students) • Some people had other stuff going on in the first two weeks • We focussed on intermediate evaluations which turned out not be that important in the end • Don’t do this in the first two weeks of Systems and Signals 414 lectures 33 / 35

  51. What we learned: What we would do differently 34 / 35

  52. What we learned: What we would do differently • Smaller team: Can’t have stand-ups with 10 people; maybe apply the one-pizza rule 34 / 35

Recommend


More recommend