引導式測驗 · 硬 等級
聲源定位和到達方向測驗
測試麥克風陣列延遲、陣列幾何形狀、迴聲、多源、角度誤差、遠場假設和身份限制。
問題 1 的 8
What physical cue can two spaced microphones provide for source direction?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- What physical cue can two spaced microphones provide for source direction?
- A direct measure of source distance
- A difference in sound arrival time
- A speaker-identity match score
- A transcript of the incoming sound
- Why is array geometry needed for DOA estimation?
- Geometry automatically reveals the source name
- Any placement produces the same delays
- Only the audio file title matters
- Timing differences must be interpreted relative to microphone positions
- A direction peak points at a wall while the speaker is elsewhere. What may explain it?
- A strong reflected arrival was mistaken for the direct sound
- Sound cannot reflect indoors
- The speaker’s words changed the array shape
- MUSIC always returns legal identities
- Which quantity is not given by a DOA angle alone?
- An angular direction relative to the array
- A candidate beam-steering direction
- The source’s exact distance from the array
- A direction to compare over time
- Why might a near source challenge a far-field DOA model?
- Near sounds cannot reach microphones
- Its wavefront may not be well approximated as planar
- Distance removes all time differences
- The speed of sound becomes zero
- What should a multi-speaker evaluation include?
- Missed and false source directions as well as angular errors
- Only one clean speaker recording
- Only the array brand
- Only a transcript WER
- What can temporal smoothing of a direction track trade?
- Guaranteed identity recognition
- More accurate distance without new data
- Removal of every room echo
- Less jitter for slower response to a moving source
- Which inference would overstep the guide’s evidence?
- “A sound arrived from the left” under the array frame
- “An echo may have changed the angle”
- “This named person spoke” from a DOA angle alone
- “The array should be calibrated”