Motion-I2V: Consistent and Controllable Image-to-Video Generation with Explicit Motion Modeling.
Xiaoyu ShiZhaoyang HuangFu-Yun WangWeikang BianDasong LiYi ZhangManyuan ZhangKa Chun CheungSimon SeeHongwei QinJifeng DaiHongsheng LiPublished in: CoRR (2024)
Keyphrases
- static images
- object motion
- temporal continuity
- space time
- visual cues
- successive frames
- image frames
- camera movement
- input video
- image motion
- image sequences
- key frames
- reference frame
- visual perception
- single image
- visual data
- single frame
- video sequences
- layered representation
- motion analysis
- camera motion
- motion estimation
- image data
- position and orientation
- normalized correlation
- motion parameters
- target object
- optical flow
- spatial and temporal
- feature points
- temporal domain
- video scene
- low level
- motion features
- input image
- motion model
- multiscale
- moving camera
- image segmentation
- moving objects
- action recognition
- image retrieval
- motion trajectories
- displacement estimation
- video surveillance
- adjacent frames
- dynamic scenes
- video content
- image features
- image classification
- video frames
- rotational motion
- video objects
- high resolution
- video signals
- ego motion
- motion capture
- ground plane
- human actions
- computer vision
- video retrieval