eduzhai > Applied Sciences > Engineering >

HRVGAN High Resolution Video Generation using Spatio-Temporal GAN

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 11 pages

Abstract: In this paper, we present a novel network for high resolution videogeneration. Our network uses ideas from Wasserstein GANs by enforcingk-Lipschitz constraint on the loss term and Conditional GANs using class labelsfor training and testing. We present Generator and Discriminator networklayerwise details along with the combined network architecture, optimizationdetails and algorithm used in this work. Our network uses a combination of twoloss terms: mean square pixel loss and an adversarial loss. The datasets usedfor training and testing our network are UCF101, Golf and Aeroplane Datasets.Using Inception Score and Fréchet Inception Distance as the evaluationmetrics, our network outperforms previous state of the art networks onunsupervised video generation.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×