WaveNet Quiz

Answers and explanations appear as you take the quiz.

1. How does WaveNet generate audio?
   • One raw sample at a time, conditioned on previous samples
   • By selecting clips from a fixed library
   • By converting spectrograms directly to text
   • All samples simultaneously with no ordering
2. What is the key convolutional innovation in WaveNet?
   • Max pooling layers
   • Dilated causal convolutions
   • Recurrent gates only
   • Fully connected layers
3. How did the original WaveNet represent quantized waveform samples for prediction?
   • As a single continuous average for the entire clip
   • As speaker IDs only
   • As a categorical distribution over 256 µ-law-quantized values
   • As a sequence of image patches
4. What was a major practical drawback of the original WaveNet?
   • It could only produce silence
   • It required no training data
   • It only worked on images
   • Sequential sampling made generation slow
5. In the speech-synthesis experiments described in the WaveNet paper, what does the model generate?
   • Raw speech waveforms sample by sample, conditioned on linguistic features
   • A mel-spectrogram that is itself the final audio
   • Only phoneme labels without audio
   • A MIDI score
6. Which organization introduced WaveNet?
   • NASA
   • DeepMind
   • OpenAI
   • Adobe
7. What does causal convolution prevent in WaveNet?
   • Using preceding samples as context
   • Generating any audio waveform
   • Using future audio samples to predict the current sample
   • Conditioning on text or speaker features
8. What activation structure does the original WaveNet use in its gated convolutional layers?
   • A ReLU output concatenated with image patches
   • A softmax over speakers only
   • A linear layer without nonlinearities
   • A tanh-filter output multiplied elementwise by a sigmoid gate