Guide-yakabatana mibvunzo · Pakati Nhanho

FastSpeech uye Non-Autoregressive TTS Quiz

Inotarisa parallel spectrogram chizvarwa, nguva yekuwedzera, FastSpeech 2 pitch uye simba mamiriro, uye inference tradeoffs.

Nzira dzekutungamira dzinoenderanaFastspeech And Non Autoregressive Tts
Mubvunzo 1 ye 8

Chii chinonzi 'chisina-autoregressive' chinorevei mumamiriro eFastSpeech?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Chii chinonzi 'chisina-autoregressive' chinorevei mumamiriro eFastSpeech?
    • Iyo inogadzira ese ma spectrogram mafuremu anoenderana kwete sequentially
    • Iyo inogadzira maodhiyo mafuremu imwe panguva mukutevedzana
    • Haikwanise kubata mitsara yakareba
    • Izvo zvinoda hapana data yekudzidziswa
  2. Nderipi dambudziko reautoregressive modhi seTacotron 2 inoita FastSpeech inogadzirisa?
    • Kusakwanisa kugadzira chero audio
    • Kunonoka kugadzira uye kusvetuka kana kudzokororwa mazwi kubva mukukanganisa kwekutarisisa
    • Kushaikwa kwerutsigiro rwechinyorwa chechirungu
    • Inoda quantum hardware
  3. Ndeipi basa reiyo 'refu regulator' muFastSpeech?
    • Inodzvanya iyo yekupedzisira audio faira
    • Inobvisa kunyarara kubva pane zvakarekodhwa
    • Inodzokorora kumiririrwa kwefonime imwe neimwe zvinoenderana neyakafanotaurwa nguva yekufananidza kureba kwekuteerera
    • Inoturikira phoneme kuita zvinyorwa
  4. Which additional speech variations does FastSpeech 2 condition on during training and inference?
    • Only speaker age and recording location
    • Video frames and subtitle color
    • A single fixed global loudness value
    • Pitch, energy, and duration information
  5. Nei FastSpeech's inference nguva isingakura nekureba kwemutsara?
    • Mafuremu ese anogadzirwa anoenderana kwete achitevedzana
    • Inofuratira mazhinji emavara ekuisa
    • Inongoburitsa zvimedu zvesekondi imwe chete
    • Inoshandisa tafura yekutarisa yemitsara yakarekodhwa
  6. How does the length regulator use predicted durations in FastSpeech?
    • It selects the final audio codec
    • It expands phoneme-level representations to align with the target spectrogram length
    • It estimates the speaker’s identity from video
    • It converts text into a sequence of labels for translation
  7. What is one tradeoff of non-autoregressive spectrogram generation?
    • It guarantees identical prosody to an autoregressive model
    • It removes the need for a waveform vocoder
    • Parallel output can be faster, but duration and speech variation must be modeled explicitly
    • It cannot accept text as input
  8. What training target does FastSpeech 2 use instead of FastSpeech’s teacher-generated simplified outputs?
    • Only predicted durations with no acoustic target
    • A translated text transcript
    • The output of a GAN discriminator only
    • Ground-truth target speech representations