Maswali yanayounganishwa na mwongozo · Kati Kiwango

Maswali ya Kutenganisha Shina la Spleeter

Jaribu uelewa wako wa miundo ya kutenganisha chanzo cha Spleeter, vinyago vya kutazama, usanidi wa shina na madai ya kasi iliyopimwa.

Njia za mwongozo zinazohusianaMgawanyiko wa Shina la Spleeter
Swali 1 ya 8

Ni kampuni gani ilitoa Spleeter awali?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  1. Ni kampuni gani ilitoa Spleeter awali?
    • Deezer
    • Spotify
    • Apple
    • Google
  2. Je, mtindo wa shina 4 wa Spleeter hutenganisha sauti kuwa nini?
    • Nyimbo, gitaa, piano, nyuzi
    • Sauti, ngoma, bass, nyingine
    • Kuongoza, maelewano, melody, rhythm
    • Kushoto, kulia, katikati, ambience
  3. Je, Spleeter hutumia usanifu gani wa mtandao wa neva?
    • LSTM ya kawaida kwenye sampuli ghafi
    • Mkusanyiko wa mti wa maamuzi
    • U-Net inayofanya kazi kwenye spectrogram
    • Mfano wa kueneza
  4. Je, Spleeter huchotaje chanzo kimoja kutoka kwa mchanganyiko?
    • Kwa kuondoa kiolezo kisichobadilika
    • Kwa kughairi awamu ya chaneli mbili
    • Kwa kuchuja na kusawazisha fasta
    • Kwa kutabiri mask laini inayotumiwa kwenye spectrogram
  5. Chanzo cha kawaida cha mabaki katika matokeo ya Spleeter ni kwamba:
    • Hutumia tena awamu ya mchanganyiko asilia kwa kila shina
    • Hutumia MP3 iliyopotea ndani
    • Sampuli za chini hadi 8 kHz
    • Inaendeshwa kwenye CPU pekee
  6. What did Deezer report about the speed of Spleeter’s pretrained 4-stem model?
    • Real-time operation with no buffering on every device
    • About 100 times faster than real time on one GPU, with buffering still needed
    • A fixed speed of 100 times real time on a phone CPU
    • It can separate audio without processing the input
  7. Which stem does Spleeter’s 5-stem configuration add to the 4-stem set?
    • Violin
    • Speech transcript
    • Piano
    • Room reverb
  8. What do Spleeter’s pretrained U-Nets estimate for the target sources?
    • A text transcript of the song
    • A fixed equalizer curve for the whole track
    • A new recording of each instrument
    • Soft masks over a spectrogram representation