MCVD: Masked Conditional Video Diffusion for Prediction, Generation, and Interpolation.
Vikram VoletiAlexia Jolicoeur-MartineauChristopher PalPublished in: CoRR (2022)
Keyphrases
- prediction accuracy
- video sequences
- video content
- video data
- video streams
- prediction algorithm
- real time
- multimedia
- video frames
- video analysis
- diffusion process
- video clips
- generation process
- video segmentation
- anisotropic diffusion
- prediction model
- probabilistic model
- video retrieval
- key frames
- video database
- scalable video coding
- spatio temporal
- dynamic scenes
- temporal information
- multimedia data
- digital video
- human activities
- video processing
- image interpolation
- real time video