# ModeRNN: Harnessing Spatiotemporal Mode Collapse in Unsupervised Predictive Learning

arXiv: https://arxiv.org/abs/2110.03882 · HTML: https://mlsanity.com/paper/2110.03882/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 43 of its 58 references that are in the map.
- First posted on arXiv 8 Oct 2021.

## Builds on (most cited first, 30 of 43)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Xception: Deep Learning with Depthwise Separable Convolutions](https://mlsanity.com/paper/1610.02357/) (2016)
- [Conditional Generative Adversarial Nets](https://mlsanity.com/paper/1411.1784/) (2014)
- [Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting](https://mlsanity.com/paper/1506.04214/) (2015)
- [Towards Accurate Generative Models of Video: A New Metric & Challenges](https://mlsanity.com/paper/1812.01717/) (2018)
- [Object-Centric Learning with Slot Attention](https://mlsanity.com/paper/2006.15055/) (2020)
- [MoCoGAN: Decomposing Motion and Content for Video Generation](https://mlsanity.com/paper/1707.04993/) (2017)
- [Unsupervised Learning of Video Representations using LSTMs](https://mlsanity.com/paper/1502.04681/) (2015)
- [Video-to-Video Synthesis](https://mlsanity.com/paper/1808.06601/) (2018)
- [Generating Videos with Scene Dynamics](https://mlsanity.com/paper/1609.02612/) (2016)
- [Dynamic Filter Networks](https://mlsanity.com/paper/1605.09673/) (2016)
- [Deep multi-scale video prediction beyond mean square error](https://mlsanity.com/paper/1511.05440/) (2015)
- [RoboNet: Large-Scale Multi-Robot Learning](https://mlsanity.com/paper/1910.11215/) (2019)
- [Stochastic Video Generation with a Learned Prior](https://mlsanity.com/paper/1802.07687/) (2018)
- [Stochastic Variational Video Prediction](https://mlsanity.com/paper/1710.11252/) (2017)
- [Recurrent Independent Mechanisms](https://mlsanity.com/paper/1909.10893/) (2019)
- [Disentangling Physical Dynamics from Unknown Factors for Unsupervised Video Prediction](https://mlsanity.com/paper/2003.01460/) (2020)
- [Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model](https://mlsanity.com/paper/1706.03458/) (2017)
- [PredRNN: A Recurrent Neural Network for Spatiotemporal Predictive Learning](https://mlsanity.com/paper/2103.09504/) (2021)
- [Stochastic Adversarial Video Prediction](https://mlsanity.com/paper/1804.01523/) (2018)
- [Action-Conditional Video Prediction using Deep Networks in Atari Games](https://mlsanity.com/paper/1507.08750/) (2015)
- [Unsupervised Learning of Disentangled Representations from Video](https://mlsanity.com/paper/1705.10915/) (2017)
- [Learning to Decompose and Disentangle Representations for Video Prediction](https://mlsanity.com/paper/1806.04166/) (2018)
- [A Review on Deep Learning Techniques for Video Prediction](https://mlsanity.com/paper/2004.05214/) (2020)
- [Video Pixel Networks](https://mlsanity.com/paper/1610.00527/) (2016)

## Built on by (most cited first, 4 of 4)

- [SpikingJelly: An open-source machine learning infrastructure platform for spike-based intelligence](https://mlsanity.com/paper/2310.16620/) (2023)
- [Deep learning for precipitation nowcasting: A survey from the perspective of time series forecasting](https://mlsanity.com/paper/2406.04867/) (2024)
- [SSLChange: A Self-supervised Change Detection Framework Based on Domain Adaptation](https://mlsanity.com/paper/2405.18224/) (2024)
- [ComS2T: A complementary spatiotemporal learning system for data-adaptive model evolution](https://mlsanity.com/paper/2403.01738/) (2024)

## Abstract (arXiv)

Learning predictive models for unlabeled spatiotemporal data is challenging in part because visual dynamics can be highly entangled in real scenes, making existing approaches prone to overfit partial modes of physical processes while neglecting to reason about others. We name this phenomenon spatiotemporal mode collapse and explore it for the first time in predictive learning. The key is to provide the model with a strong inductive bias to discover the compositional structures of latent modes. To this end, we propose ModeRNN, which introduces a novel method to learn structured hidden representations between recurrent states. The core idea of this framework is to first extract various components of visual dynamics using a set of spatiotemporal slots with independent parameters. Considering that multiple space-time patterns may co-exist in a sequence, we leverage learnable importance weights to adaptively aggregate slot features into a unified hidden representation, which is then used to update the recurrent states. Across the entire dataset, different modes result in different responses on the mixtures of slots, which enhances the ability of ModeRNN to build structured representations and thus prevents the so-called mode collapse. Unlike existing models, ModeRNN is shown to prevent spatiotemporal mode collapse and further benefit from learning mixed visual dynamics.