D3: Recognizing dynamic scenes with deep dual descriptor based on key frames and key segments.
Sungeun HongJongbin RyuWoobin ImHyun Seung YangPublished in: Neurocomputing (2018)
Keyphrases
- dynamic scenes
- key frames
- video sequences
- space time
- multi view
- moving objects
- video data
- moving camera
- static scenes
- video segments
- video retrieval
- video content
- background subtraction
- motion segmentation
- video summarization
- video shots
- instructional videos
- feature vectors
- video frames
- video clips
- object recognition
- multiple views
- motion vectors
- visual features
- frame rate
- semi supervised
- video segmentation
- motion estimation
- computer vision
- object tracking
- image quality
- optical flow