LAtt-Yolov8-seg: Video Real-time Instance Segmentation for Urban Street Scenes Based on Focused Linear Attention Mechanism.
Xinqi ZhangTuo DongLiqi YanZhenglei YangJianhui ZhangPublished in: CVDL (2024)
Keyphrases
- real time
- street scenes
- semantic segmentation
- attention mechanism
- surveillance videos
- video data
- conditional random fields
- level set
- superpixels
- segmentation algorithm
- video content
- object categories
- video sequences
- segmentation method
- saliency map
- weakly supervised
- multiscale
- video frames
- image segmentation
- visual attention
- image processing
- scene classification
- object classes
- shape prior
- vision system
- multiple objects
- key frames
- event detection
- graph cuts
- image retrieval
- pairwise