Abstract: Most existing driver attention prediction models rely on 2D or 3D convolutional neural networks for static or local temporal modeling, which limits their ability to capture long-range ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results