Decomposing Motion and Content for Natural Video Sequence Prediction

25 Jun 2017  ·  Ruben Villegas, Jimei Yang, Seunghoon Hong, Xunyu Lin, Honglak Lee ·

We propose a deep neural network for the prediction of future frames in natural video sequences. To effectively handle complex evolution of pixels in videos, we propose to decompose the motion and content, two key components generating dynamics in videos. Our model is built upon the Encoder-Decoder Convolutional Neural Network and Convolutional LSTM for pixel-level prediction, which independently capture the spatial layout of an image and the corresponding temporal dynamics. By independently modeling motion and content, predicting the next frame reduces to converting the extracted content features into the next frame content by the identified motion features, which simplifies the task of prediction. Our model is end-to-end trainable over multiple time steps, and naturally learns to decompose motion and content without separate training. We evaluate the proposed network architecture on human activity videos using KTH, Weizmann action, and UCF-101 datasets. We show state-of-the-art performance in comparison to recent approaches. To the best of our knowledge, this is the first end-to-end trainable network architecture with motion and content separation to model the spatiotemporal dynamics for pixel-level future prediction in natural videos.

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 Benchmark
Video Prediction KTH MCnet + Residual PSNR 26.29 # 20
SSIM 0.806 # 17
Cond 10 # 1
Pred 20 # 1

Results from Other Papers

Task Dataset Model Metric Name Metric Value Rank Source Paper Compare
Video Prediction KTH MCnet PSNR 25.95 # 23
SSIM 0.804 # 19
Cond 10 # 1
Pred 20 # 1