Predicting Scene Parsing and Motion Dynamics in the Future.
Xiaojie JinHuaxin XiaoXiaohui ShenJimei YangZhe LinYunpeng ChenZequn JieJiashi FengShuicheng YanPublished in: CoRR (2017)
Keyphrases
- image sequences
- moving objects
- dynamic scenes
- image motion
- object motion
- fixed camera
- predicting future
- camera motion
- moving camera
- multiple moving objects
- static scenes
- d scene
- optical flow fields
- motion estimation
- vision sensors
- motion analysis
- motion features
- consecutive frames
- optical flow
- video sequences
- camera images
- static images
- three dimensional
- space time
- visual scene
- motion model
- uncalibrated cameras
- relative position
- point features
- moving observer
- visual data
- motion segmentation
- ground plane
- natural language processing
- input video
- apparent motion
- motion parameters
- video scene
- motion detection
- relative depth
- global motion
- camera movement
- target object
- affine structure
- natural language
- single image
- human walking
- input image
- visual motion
- traffic scenes
- imaging process
- human motion
- multiple images
- ego motion
- structure from motion
- dynamical systems
- multi view
- dynamic textures
- physical constraints
- flow field
- motion patterns
- layered representation