MoCoGAN: Decomposing Motion and Content for Video Generation.
Sergey TulyakovMing-Yu LiuXiaodong YangJan KautzPublished in: CVPR (2018)
Keyphrases
- video footage
- space time
- multimedia
- input video
- moving camera
- video sequences
- video data
- online video
- key frames
- object motion
- image sequences
- temporal filtering
- generation process
- static images
- video content
- successive frames
- low frame rate
- spatial and temporal
- temporal continuity
- visual cues
- motion features
- video frames
- motion analysis
- video representation
- motion estimation
- motion model
- real time
- surveillance videos
- motion trajectories
- shot change detection
- visual data
- multimedia data
- dynamic scenes
- temporal domain
- reference frame
- video database
- video material
- camera motion
- motion parameters
- video segments
- video streams
- multimedia documents
- optical flow
- layered representation
- moving objects
- single frame
- visual information
- periodic motion
- spatio temporal
- lecture videos
- motion capture data
- temporal consistency
- temporal coherence
- human motion
- video surveillance
- user generated
- image frames
- dynamic textures
- humanoid robot
- motion field
- video scene
- video summarization
- motion segmentation
- feature points
- motion detection
- video shots
- video analysis