LMVP is a GAN-based video prediction model that introduces a motion guider component and information leakage from the discriminator to achieve claimed state-of-the-art results on synthetic and real videos without labeled data.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LMVP: Video Predictor with Leaked Motion Information
LMVP is a GAN-based video prediction model that introduces a motion guider component and information leakage from the discriminator to achieve claimed state-of-the-art results on synthetic and real videos without labeled data.