将视觉场景图转换为图像说明（英文版）

英文标题：Transforming Visual Scene Graphs to Image Captions中文摘要：本论文提出了一种新的图像 / 视频描述方法，称之为 TSG，它使用多头注意力机制 (MHA) 和混合专家解码器，将场景图转换为更具描述性的字幕，并在 MS-COCO 数据集上取得了很好的效果。英文摘要：We propose to Transform Scene Graphs (TSG) into more descriptive captions. InTSG, we apply multi-head attention (MHA) to design the Graph Neural Network(GNN) for embedding scene graphs. After embedding, different graph embeddingscontain diverse specific knowledge for generating the words with differentpart-of-speech, e.g., object/attribu