PredRNN: A Recurrent Neural Network for Spatiotemporal Predictive Learning

17 Mar 2021  ·  Yunbo Wang, Haixu Wu, Jianjin Zhang, Zhifeng Gao, Jianmin Wang, Philip S. Yu, Mingsheng Long ·

The predictive learning of spatiotemporal sequences aims to generate future images by learning from the historical context, where the visual dynamics are believed to have modular structures that can be learned with compositional subsystems. This paper models these structures by presenting PredRNN, a new recurrent network, in which a pair of memory cells are explicitly decoupled, operate in nearly independent transition manners, and finally form unified representations of the complex environment. Concretely, besides the original memory cell of LSTM, this network is featured by a zigzag memory flow that propagates in both bottom-up and top-down directions across all layers, enabling the learned visual dynamics at different levels of RNNs to communicate. It also leverages a memory decoupling loss to keep the memory cells from learning redundant features. We further propose a new curriculum learning strategy to force PredRNN to learn long-term dynamics from context frames, which can be generalized to most sequence-to-sequence models. We provide detailed ablation studies to verify the effectiveness of each component. Our approach is shown to obtain highly competitive results on five datasets for both action-free and action-conditioned predictive learning scenarios.

PDF Abstract

Results from the Paper


 Ranked #1 on Video Prediction on KTH (Cond metric)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Video Prediction KTH PredRNN-V2 LPIPS 0.139 # 11
PSNR 28.37 # 7
SSIM 0.839 # 12
Cond 10 # 1
Pred 20 # 1
Video Prediction Moving MNIST PredRNN-V2 MSE 48.4 # 26
SSIM 0.891 # 21
LPIPS 0.071 # 3
Weather Forecasting SEVIR PredRNN MSE 3.9014 # 4
mCSI 0.4080 # 4

Methods