Towards Reversible De-Identification in Video Sequences Using 3D Avatars and Steganography.
Martin BlazevicKarla BrkicTomislav HrkacPublished in: CoRR (2015)
Keyphrases
- video sequences
- virtual world
- virtual environment
- information hiding
- markov chain
- object tracking
- moving objects
- visual tracking
- cellular automata
- image sequences
- video data
- video segmentation
- video camera
- frame rate
- motion estimation
- virtual reality
- camera motion
- motion vectors
- human motion
- dynamic scenes
- data hiding
- high capacity
- space time
- facial expressions
- inter frame
- secret information
- hidden information
- spread spectrum
- multimedia
- video frames
- moving camera
- jpeg images
- human actions
- video content
- appearance model
- computer vision