Guide-yakabatana mibvunzo · Pakati Nhanho

DiffWave Diffusion Vocoder Quiz

Dzidza kuti DiffWave inofanotaura sei ruzha rwekugadzira waveforms, mamiriro pane acoustic maficha, uye zviyero zvesampling kumhanya kupesana nemhando.

Nzira dzekutungamira dzinoenderanaDiffwave Diffusion Vocoder
Mubvunzo 1 ye 8

Ko DiffWave inogadzira sei odhiyo panguva yekufungidzira?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Ko DiffWave inogadzira sei odhiyo panguva yekufungidzira?
    • Nekudzokorodza kudzokorodza ruzha rusina kurongeka kuita waveform
    • Nekuenzanisira imwe waveform kukosha panguva autoregressively
    • Nekubatanidza mayuniti ekutaura akarekodhwa
    • Nekumhanya imwechete GAN yekupfuura
  2. Chii chinonzi DiffWave network chinonyatso dzidza kufanotaura panguva yekudzidziswa?
    • Iyo yakachena waveform yakananga mune imwe pfuti
    • Ruzha rwakawedzerwa padanho rakapihwa rekupararira
    • The text transcript
    • Kuzivikanwa kwemutauri
  3. Chii chinonzi DiffWave's main practical drawback ichienzaniswa neGAN vocoders?
    • Haikwanise kumira pane spectrograms
    • Inoshanda chete pamitengo yakaderera yemuenzaniso
    • Inogona kuda matanho akawanda ekuenzanisa, zvichiita kuti iite zvishoma
    • Zvinoda kudzidziswa kweanopikisa kuti zvigadzikane
  4. Ndeupi mukana une DiffWave pamusoro peGAN vocoder mukudzidziswa?
    • Izvo hazvidi neural network
    • Izvo hazvidi data
    • Inoburitsa odhiyo ipapo munhanho imwe
    • Inodzivisa kudzidziswa kwevadzivisi kusina kugadzikana
  5. Ndechipi chivakwa chinofanana neDiffWave's noise-prediction network?
    • A WaveNet-style yakawedzera convolutional network
    • Idzokororo yemutauro weLSTM
    • A vision transformer
    • A simple linear regression
  6. Zvichida kuti mashoma sei matanho ekuenzanisa anogona kunyatso kurongeka DiffWave masheti achichengeta kunaka kwakanaka?
    • Chaiyo 1000 nhanho nguva dzose
    • Anenge matanho matanhatu
    • Around 500 nhanho
    • Nhanho imwe chete isina tuning
  7. What acoustic representation can condition DiffWave when it is used as a vocoder?
    • A list of subtitle line numbers
    • A CT scan of an ancient scroll
    • A mel-spectrogram
    • A source-code dependency graph
  8. What loss does DiffWave use to train its noise-prediction network?
    • A word-error-rate objective
    • A discriminator feature-matching loss only
    • A cosine-similarity loss between image embeddings
    • A weighted L2 loss on the predicted noise