Wavenet and deep learning - based TTS systems

123456789101112131415161718192021222324252627
Across
  1. 6. Neural architecture using self-attention
  2. 7. RNN variant handling long-term dependencies
  3. 8. Field that enables machines to mimic human intelligence
  4. 9. Simplified recurrent neural network
  5. 12. Converts text into hidden representations
  6. 14. Mechanism to focus on relevant input parts
  7. 16. Converts acoustic features into waveform audio
  8. 17. Frequency-based audio representation for TTS
  9. 18. Predicts future samples from past outputs
  10. 20. Computing system inspired by biological neurons
  11. 23. Network effective in feature extraction
  12. 25. Converts encoded features into speech output
  13. 27. Artificial production of human speech
Down
  1. 1. Raw audio signal representation
  2. 2. Generating speech from trained model
  3. 3. Delay between input and speech output
  4. 4. Process of learning model parameters
  5. 5. How human-like synthesized speech sounds
  6. 10. Machine learning using multi-layer neural networks
  7. 11. Collection of text and audio samples
  8. 13. Converts written text into natural-sounding speech
  9. 15. Written representation of a sound
  10. 19. Intonation, stress, and rhythm in speech
  11. 21. Network designed for sequence modeling
  12. 22. Single system trained from text to speech
  13. 23. Large structured speech dataset
  14. 24. Basic sound unit of spoken language
  15. 26. DeepMind model that generates speech one sample at a time