By Qiang Wang, Jin Gao, Junliang Xing, Mengdan Zhang, Weiming Hu
Discriminant Correlation Filters (DCF) based methods now become a kind of dominant approach to online object tracking. The features used in these methods, however, are either based on hand-crafted features like HoGs, or convolutional features trained independently from other tasks like image classification. In this work, we present an end-to-end lightweight network architecture, namely DCFNet, to learn the convolutional features and perform the correlation tracking process simultaneously.
Requirements for MatConvNet 1.0-beta23(see: MatConvNet)
- Downloading MatConvNet
git clone https://github.com/vlfeat/matconvnet.git
- Compiling MatConvNet
Run the following command from the MATLAB command window:
run <matconvnet>/matlab/vl_compilenn
git clone --depth=1 https://github.com/foolwood/DCFNet.git
The file demo/demoDCFNet.m
is used to test our algorithm.
To verify OTB and VOT performance, you can simple copy DCFNet/
into OTB toolkit and integrate track4vot/
to VOT toolkit.
1.Download the training data.
TColor-128:[LINK]
UAV123: [GoogleDrive]
NUS_PRO:[GoogleDrive] (part1)(part2)]
It should have this basic structure
data
|-- NUS_PRO
|-- Temple-color-128
|-- UAV123
2.Run training/train_cnn_dcf.m
to train a model.
You can choose the network architecture by setting opts.networkType = 21
(This parameter is 21 by default)
AUC on OTB2013 and OTB2015(OPE)
VOT2015 EAO result
If you find DCFNet useful in your research, please consider citing:
@article{wang17dcfnet,
Author = {Qiang Wang, Jin Gao, Junliang Xing, Mengdan Zhang, Weiming Hu},
Title = {DCFNet: Discriminant Correlation Filters Network for Visual Tracking},
Journal = {arXiv preprint arXiv:1704.04057},
Year = {2017}
}