A High-Dimensional Video Sequence Completion Method with Traffic Data Completion Generative Adversarial Networks

Figure 5

The completion results of multiple frames on Caltech pedestrian dataset. (a) Input. (b) Details. (c) Output.