Adaptive convolutional layer selection based on historical retrospect for visual tracking

Adaptive convolutional layer selection based on historical retrospect for visual tracking

Visual tracking has recently gained a great advance with the use of the convolutional neural network (CNN). Usually, existing CNN‐based trackers exploit the features from a single layer or a certain combination of multiple layers. However, these features only characterise an object from an invariabl...

Celý popis

Podrobná bibliografie
Hlavní autoři: Fuhui Tang, Xiankai Lu, Xiaoyu Zhang, Lingkun Luo, Shiqiang Hu, Huanlong Zhang
Médium: Článek
Jazyk:English
Vydáno: Wiley 2019-04-01
Edice:IET Computer Vision
Témata:
invariable aspect
scene variation
convolutional layer selection method
robust appearance representation
different convolutional layers
tracking prediction
On-line přístup:https://doi.org/10.1049/iet-cvi.2018.5194

Internet

https://doi.org/10.1049/iet-cvi.2018.5194

Podobné jednotky