PredRNN: A Recurrent Neural Network for Spatiotemporal Predictive Learning
This work addresses the challenge of generating future images from historical sequences, which is important for applications like video prediction, but it is incremental as it builds on existing LSTM-based methods.
The paper tackles the problem of spatiotemporal predictive learning by introducing PredRNN, a recurrent neural network that decouples memory cells and uses a zigzag memory flow to improve communication across layers, achieving competitive results on five datasets.
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.