Video Captioning with Aggregated Features Based on Dual Graphs and Gated Fusion