PESQ and STOI Speech Quality Metrics Quiz

Answers and explanations appear as you take the quiz.

1. What does PESQ estimate when comparing a degraded speech signal with a reference?
   • Perceived speech quality
   • The transcript’s word error rate
   • The speaker’s identity
   • The network’s packet latency
2. What does STOI primarily predict?
   • Audio bitrate
   • Speech intelligibility
   • Speaker identity
   • Reverberation time
3. Both PESQ and STOI are described as 'intrusive' metrics. What does that mean?
   • They damage the audio file
   • They can only run in real time
   • They require a clean reference signal for comparison
   • They need human listeners
4. What is the commonly reported output range for STOI?
   • -0.5 to 4.5
   • 1 to 100
   • 0 to 10
   • 0 to 1
5. PESQ models human hearing using which kind of perceptual frequency scale?
   • Bark/loudness bands
   • Linear Hz spacing
   • MIDI note numbers
   • Decibel-only weighting
6. Why are non-intrusive metrics like DNSMOS attracting interest as alternatives?
   • They produce lower scores
   • They need no clean reference signal
   • They only work for music
   • They are part of ITU-T P.862
7. What use case did the original ITU-T P.862 Recommendation describe for PESQ?
   • Identifying speakers in music recordings
   • Transcribing wideband speech into text
   • Predicting subjective quality of narrow-band telephone speech and codecs
   • Measuring network packet delay
8. What is the current ITU status of Recommendation P.862 (PESQ)?
   • An active current recommendation for all speech domains
   • A required open-source software package
   • A standard only for speaker diarization
   • Superseded and withdrawn; the recommendation was deleted