CMS-LSTM: Context-Embedding and Multi-Scale Spatiotemporal-Expression LSTM for Video Prediction.
Zenghao ChaiChun YuanZhihui LinYunpeng BaiPublished in: CoRR (2021)
Keyphrases
- multiscale
- recurrent neural networks
- spatial and temporal
- space time
- prediction accuracy
- multimedia
- video sequences
- video frames
- real time
- vector space
- video data
- digital video
- prediction algorithm
- spatio temporal
- image processing
- prediction error
- video analysis
- spatiotemporal features
- video streams
- scale space
- multiresolution
- video clips
- contextual information
- natural images
- neural network