• il y a 14 ans
On the left you have the video, on the right the bottom-up saliency map. It uses only motion features (speed and direction), no static features.