Detection and tracking based tubelet generation for video object detection
Lists
Only you see this
Abstract
Highlights•Precise location and temporal consistency can be guaranteed with the DAT framework.•Serializing detection proposals with optical flow can generate precise tubelets.•Identifying each object with the temporal information of tracking trajectory.•Tubelet fusion produces better results by integrating union/concatenation operation.
BibTeX
@article{wang2019detection,
author = {Bin Wang and Sheng Tang and Jun Bin Xiao and Quan-Feng Yan and Yong-Dong Zhang},
title = {{Detection and tracking based tubelet generation for video object detection}},
journal = {J. Vis. Commun. Image Represent.},
year = {2019},
url = {https://openreview.net/forum?id=oz8lZXE86O},
}Loading related content...