Exposing AI-generated videos with motion magnification.
Jianwei FeiZhihua XiaPeipeng YuFengjun XiaoPublished in: Multim. Tools Appl. (2021)
Keyphrases
- periodic motion
- image sequences
- moving camera
- key frames
- video sequences
- motion features
- motion estimation
- motion analysis
- motion model
- recognition of human actions
- optical flow
- artificial intelligence
- dynamic scenes
- low frame rate
- human motion
- intelligent systems
- super resolution
- high resolution
- recognizing human actions
- spatial and temporal
- temporal coherence
- static images
- video objects
- space time
- successive frames
- surveillance videos
- video frames
- camera motion
- input video
- video representation
- motion cues
- video stabilization
- video data
- expert systems
- spatial resolution
- motion detection
- case based reasoning
- dynamic textures
- visual data
- video content
- video summarization
- knowledge representation
- motion parameters
- video surveillance
- spatio temporal
- global motion
- spatial and temporal information
- rigidly moving
- adjacent frames
- temporal domain
- point correspondences
- video retrieval
- feature points
- object motion
- motion field
- video shots
- flow field