Idanwo ti o sopọ mọ itọsọna · Alabọde Ipele

Idanwo Iṣẹlẹ Ohun

Study sound classes and time boundaries, with polyphony, weak labels, the DCASE baseline, onset tolerance, and missing-class annotations.

Jẹmọ itọsọna awọn ọnaOhun iṣẹlẹ erin
Ibeere 1 ti 8

What extra output does temporal sound event detection estimate beyond a clip-level tag?

Awotẹlẹ gbogbo ibeere8 ibeere

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

  1. What extra output does temporal sound event detection estimate beyond a clip-level tag?
    • Only the audio file’s compression ratio
    • Only its filename’s alphabetical order
    • Event onset and offset boundaries within the clip
    • Only the display color of a waveform
  2. Kini idi ti wiwa iṣẹlẹ ohun nigbagbogbo ṣe apejuwe bi iṣoro polyphonic kan?
    • Nitoripe o mu awọn akọsilẹ orin nikan
    • Nitoripe ọpọlọpọ awọn ohun le ni lqkan ati waye ni nigbakannaa
    • Nitoripe o nilo awọn gbohungbohun pupọ
    • Nitori ohun ti wa ni nigbagbogbo sitẹrio
  3. Kini 'iṣamisi alailagbara' tumọ si ni data ikẹkọ SED?
    • Awọn aami ti wa ni ìpàrokò
    • Didara ohun ko dara
    • Awọn ohun idakẹjẹ nikan ni aami
    • Awọn agekuru jẹ aami pe iṣẹlẹ kan ṣẹlẹ ṣugbọn kii ṣe deede nigbati
  4. Which classifier structure is identified in the cited DCASE 2024 baseline?
    • A convolutional recurrent neural network with added pretrained frame features
    • Only a static spreadsheet sort
    • Only an unchanged relational database table
    • Only a mandatory image-file compressor
  5. Which time-frequency representation is used in the documented sound-detection baseline context described here?
    • Only a speech transcript
    • A log-mel representation
    • Only a JPEG photograph
    • Only an unchanged output filename
  6. Which cited research challenge supplies the 2018 and 2024 sound-event task settings discussed here?
    • Awọn ere Olympic
    • A retail shopping holiday
    • A mandatory social-network ranking
    • DCASE
  7. Considering only the DCASE 2018 onset collar of 200 ms, does a 1.15 s prediction satisfy it for a 1.00 s reference onset?
    • Yes; the onset error is 150 ms
    • No; the error is 1.15 s
    • No; every predicted onset must be numerically identical
    • Yes; onset errors are never evaluated
  8. In the heterogeneous 2024 task, how does the baseline handle outputs for classes whose annotations are missing in a training dataset?
    • Treats every unannotated class as certainly absent
    • Copies all missing labels from the final test answers
    • Masks those unavailable outputs in the relevant loss computation
    • Eliminates every labeled class from the model