Motion Prediction Under Multimodality with Conditional Stochastic Networks

Abstract

Given a visual history, multiple future outcomes for a video scene are equally probable, in other words, the distribution of future outcomes has multiple modes. Multimodality is notoriously hard to handle by standard regressors or classifiers: the former regress to the mean and the latter discretize a continuous high dimensional output space. In this work… (More)

Topics

8 Figures and Tables

Cite this paper

@article{Fragkiadaki2017MotionPU, title={Motion Prediction Under Multimodality with Conditional Stochastic Networks}, author={Katerina Fragkiadaki and Jonathan Huang and Alex Alemi and Sudheendra Vijayanarasimhan and Susanna Ricco and Rahul Sukthankar}, journal={CoRR}, year={2017}, volume={abs/1705.02082} }