YOLOv4-dense: A smaller and faster YOLOv4 for real-time edge-device based object detection in traffic scene.
Yue JiangWenjing LiJun ZhangFang LiZhongcheng WuPublished in: IET Image Process. (2023)
Keyphrases
- real time
- object detection
- scene understanding
- scene recognition
- vision system
- scene flow
- video sequences
- live video
- high frame rate
- video rate
- d scene
- optical flow fields
- object recognition
- video images
- edge information
- three dimensional
- real time traffic
- scene classification
- object categories
- face detection
- camera tracking
- edge detection
- data acquisition
- sift flow
- object segmentation
- densely sampled
- control system
- multi class
- traffic conditions
- dynamic scenes
- traffic monitoring
- multi view
- network traffic
- object detectors
- pedestrian detection
- ground plane