Aggregating Motion and Attention for Video Object Detection.
Ruyi ZhangZhenjiang MiaoCong MaShanshan HaoPublished in: ACPR (2) (2019)
Keyphrases
- object detection
- space time
- object motion
- spatial and temporal
- key frames
- video sequences
- input video
- temporal filtering
- video data
- motion analysis
- motion features
- static images
- dynamic scenes
- successive frames
- video frames
- surveillance videos
- low frame rate
- moving camera
- motion estimation
- video streams
- body motions
- visual data
- video footage
- image sequences
- action detection
- temporal consistency
- video content
- image frames
- visual cues
- layered representation
- video objects
- motion trajectories
- temporal coherence
- spatio temporal
- shot change detection
- multimedia
- computer vision
- motion field
- motion model
- camera motion
- temporal continuity
- motion detection
- video analysis
- background subtraction
- video summarization
- optical flow
- single frame
- pedestrian detection
- face detection
- motion vectors
- video clips
- human motion
- wide area motion imagery
- dynamic textures
- multi class
- camera movement
- object categories
- motion capture data
- temporal structure
- video retrieval
- reference frame
- motion segmentation
- visual attention
- object segmentation
- visual features
- scene understanding
- moving objects
- object recognition
- image motion