Idanwo ti o sopọ mọ itọsọna · Alabọde Ipele

MelGAN Generative Vocoder adanwo

Understand how MelGAN maps mel-spectrograms to waveforms, uses convolutional upsampling, and trains discriminators at multiple scales.

Jẹmọ itọsọna awọn ọnaMelgan Generative Vocoder
Ibeere 1 ti 8

Iṣawọle wo ni MelGAN ṣe iyipada si ọna igbi ohun aise kan?

Awotẹlẹ gbogbo ibeere8 ibeere

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

  1. Iṣawọle wo ni MelGAN ṣe iyipada si ọna igbi ohun aise kan?
    • A mel-spectrogram
    • Aise ọrọ kikọ
    • Faili MIDI kan
    • Lattice phoneme kan
  2. Kini idi ti MelGAN yiyara pupọ ju WaveNet lọ ni itọkasi?
    • O nlo iwọn ayẹwo ti o kere ju
    • Kii ṣe autoregressive ati pe o ṣe agbejade ohun ni iwe-iwọle ti o jọra kan
    • O gbalaye nikan lori TPUs
    • O fo awọn ẹgbẹ igbohunsafẹfẹ giga
  3. Apẹrẹ iyasọtọ iyasọtọ wo ni MelGAN ṣafihan?
    • A nikan agbaye discriminator
    • A iyasoto-orisun transformer
    • Iyatọ-ọpọlọpọ ti n ṣiṣẹ lori ohun ni awọn ipinnu pupọ
    • Ko si iyasoto rara
  4. Ipadanu wo ni o ṣe iranlọwọ fun iduroṣinṣin ikẹkọ ọta MelGAN?
    • Cross-entropy lori awọn ami ọrọ
    • Itumọ ero Dimegilio pipadanu
    • Pipadanu ala Triplet
    • Pipadanu ibaamu ẹya-ara laarin awọn iṣẹ ṣiṣe iyasoto
  5. Bawo ni olupilẹṣẹ MelGAN ṣe alekun aaye gbigba rẹ?
    • Nipa lilo awọn convolutions ti o gbooro ni awọn bulọọki iyokù
    • Nipa fifi awọn ipele LSTM loorekoore kun
    • Nipa lilo akiyesi lori gbogbo agekuru
    • Nipa stacking pooling fẹlẹfẹlẹ
  6. What subjective evaluation measure did the MelGAN paper report?
    • Word error rate (WER)
    • Mean Opinion Score (MOS)
    • BLEU score
    • Peak signal-to-noise ratio (PSNR) only
  7. What do MelGAN’s transposed convolutions do in the generator?
    • Convert the waveform back into text
    • Choose the phonemes in the input sentence
    • Upsample the mel-spectrogram representation toward the audio waveform rate
    • Reduce the number of audio samples to zero
  8. Which audio domains did the MelGAN paper use to illustrate the generality of its techniques?
    • Machine translation, image captions, and video subtitling
    • Only speech recognition and speaker identification
    • Text summarization, code generation, and image classification
    • Speech synthesis, music-domain translation, and unconditional music synthesis