Audio Captioning Quiz

Answers and explanations appear as you take the quiz.

1. How does audio captioning differ from automatic speech recognition?
   • It only works on silence
   • It is slower but identical in output
   • It describes the overall acoustic scene rather than transcribing spoken words
   • It converts text into audio
2. Which published datasets supply general-audio descriptions for captioning in this guide?
   • MNIST and CIFAR only
   • Clotho and AudioCaps
   • A mandatory bank-transaction table
   • Only a raw network-password list
3. What architecture does the cited original Clotho baseline use to produce caption words from audio features?
   • Only a fixed filename lookup
   • Only a volume meter with no text output
   • Only an unchanged grammar rule list
   • An encoder-decoder scheme with attention
4. Why is audio captioning valuable for accessibility?
   • It describes non-speech sounds for deaf and hard-of-hearing users
   • It increases audio volume
   • It deletes background noise
   • It encrypts the audio
5. Which reference-comparison metric does the Clotho paper use among its caption evaluations?
   • RGB color values
   • CIDEr
   • Audio sample rate alone
   • Only a decibel reading
6. What is a major ongoing challenge in audio captioning?
   • Audio files are too small to process
   • There are no datasets available
   • Captions can only be one word long
   • Producing descriptions that are both accurate and avoid hallucinated details
7. How many reference captions per audio file does the original Clotho paper report?
   • Five
   • One
   • Fifteen
   • One hundred
8. How many log-mel-band features form each input frame in the original Clotho baseline described in the paper?
   • 8
   • 20
   • 64
   • 256