English

CrevNet: Conditionally Reversible Video Prediction

Computer Vision and Pattern Recognition 2019-10-28 v1 Machine Learning

Abstract

Applying resolution-preserving blocks is a common practice to maximize information preservation in video prediction, yet their high memory consumption greatly limits their application scenarios. We propose CrevNet, a Conditionally Reversible Network that uses reversible architectures to build a bijective two-way autoencoder and its complementary recurrent predictor. Our model enjoys the theoretically guaranteed property of no information loss during the feature extraction, much lower memory consumption and computational efficiency.

Keywords

Cite

@article{arxiv.1910.11577,
  title  = {CrevNet: Conditionally Reversible Video Prediction},
  author = {Wei Yu and Yichao Lu and Steve Easterbrook and Sanja Fidler},
  journal= {arXiv preprint arXiv:1910.11577},
  year   = {2019}
}
R2 v1 2026-06-23T11:54:39.229Z