Prosody Modeling Quiz

Answers and explanations appear as you take the quiz.

1. Which set of features best describes prosody in speech?
   • Pitch, rhythm, stress, and loudness
   • Vocabulary, grammar, and spelling
   • Microphone quality and sample rate
   • Word count and sentence length
2. In prosody modeling, what does the fundamental frequency (F0) represent?
   • The volume of the recording
   • The rate at which the vocal folds vibrate, perceived as pitch
   • The number of phonemes per second
   • The background noise level
3. How did FastSpeech 2 improve prosody control over earlier models?
   • It removed all pitch information to simplify training
   • It only worked with pre-recorded human prosody
   • It explicitly predicts duration, pitch, and energy as separate features
   • It replaced neural networks with rule-based punctuation parsing
4. Why is predicting prosody from text alone considered a one-to-many problem?
   • Because text is always longer than the audio
   • Because each letter maps to exactly one pitch
   • Because there is only one correct way to read any sentence
   • Because a single sentence can be spoken with many valid intonations and rhythms
5. In a corpus of 410 naturally occurring American-English yes/no questions, which nuclear contour was most frequent?
   • A low-rise contour, L*H-H%
   • Always a high plateau
   • Always a flat contour
   • A fall in every question
6. Why might a TTS system use a reference audio clip when generating speech?
   • To copy the exact words to say
   • To set a specific speaking style or prosody for the output
   • To reduce the file size of the result
   • To convert text into a different language
7. What can shifting contrastive stress change in an otherwise identical sentence?
   • The spelling of every word
   • The speaker’s vocal-fold anatomy
   • Which person, action or detail the speaker presents as the focus
   • The number of phonemes in the sentence
8. Why can punctuation alone be insufficient for choosing natural prosody?
   • Punctuation determines the exact vocal identity
   • Text contains no words or syntax
   • All sentences have one possible spoken contour
   • The intended emphasis and phrasing can depend on surrounding discourse