Figure 5
A figure showing five rows of four images, each representing a temporal sequence of a pedestrian crossing prediction task.

Some examples correctly predicted by our model. Red bounding boxes indicate pedestrian samples that cross the road, while green ones indicate samples that do not cross. The first and second rows show samples of pedestrians walking along the edge of the road. The third row shows the case of pedestrians standing on the roadside in place. The fourth and fifth rows show cases of occlusion and unclear images.

or Create an Account

Close subscription notice
Close access options