Table 7.

Convolutional neural network architectures for tracking.

Model	Highlight	Source Code (Framework)	Reference
GOTURN [223]	100 fps, feed-forward network, object motion and appearance	GitHub [224] (C++)	[75]
SlowFast network [225]	Low and high frame rates, slow and high pathways, lightweight network	GitHub [226] (PyTorch)	[47]
Two-stream CNN [221]	Complementary information on appearance, motion between frames,	GitHub [227] (Python)	[80]
(Inception, ResNet, VGG, and Xception) with LSTM [222]	Recurrent convolution, CNN, doubly deep in spatial and temporal layers, LSTM	GitHub [228] (PyTorch) GitHub [229] (PredNet)	[86,87,119], etc.

Note: CNN is convolutional neural network; GOTRUN is generic object tracking using regression networks; LSTM is long short-term memory; ResNet is residual network; and VGG is visual geometry group.