Kuis terkait panduan · Sedang Tingkat

Kuis Deteksi Peristiwa Suara

Pelajari kelas suara dan batasan waktu, dengan polifoni, label lemah, garis dasar DCASE, toleransi permulaan, dan anotasi kelas yang hilang.

Jalur panduan terkaitDeteksi Peristiwa Suara
Pertanyaan 1 dari 8

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

Pratinjau setiap pertanyaan8 pertanyaan

Jawaban dan penjelasan muncul saat Anda mengikuti kuis.

  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. Mengapa deteksi peristiwa suara sering digambarkan sebagai masalah polifonik?
    • Karena hanya menangani not balok saja
    • Karena beberapa suara bisa tumpang tindih dan terjadi secara bersamaan
    • Karena memerlukan banyak mikrofon
    • Karena audio selalu stereo
  3. Apa yang dimaksud dengan 'pelabelan lemah' dalam data pelatihan SED?
    • Labelnya dienkripsi
    • Kualitas audionya buruk
    • Hanya suara pelan yang diberi label
    • Klip diberi label bahwa suatu peristiwa terjadi tetapi tidak diberi label kapan tepatnya
  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?
    • Pertandingan Olimpiade
    • 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