Figure 3
Comparison between three different deep network architectures for action recognition. (a): Recurrent Neural Networks (e.g., LSTM); (b): convolutional networks Networks (e.g., 3D CNN); (c): two-stream convolutional networks (e.g., RGB - optical flow). Refer to the image caption for details.

Comparison between three different deep network architectures for action recognition. (a): Recurrent Neural Networks (e.g., LSTM); (b): convolutional networks Networks (e.g., 3D CNN); (c): two-stream convolutional networks (e.g., RGB - optical flow).

This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal