SEE-Net improves video prediction by using frame shuffling to enforce learning of natural temporal order, reporting state-of-the-art results on three synthetic and real-world datasets.
Unsupervised learning of video representations using lstms
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
-
Order Matters: Shuffling Sequence Generation for Video Prediction
SEE-Net improves video prediction by using frame shuffling to enforce learning of natural temporal order, reporting state-of-the-art results on three synthetic and real-world datasets.