论文标题
RERTR：场景图生成的关系变压器
RelTR: Relation Transformer for Scene Graph Generation
论文作者
论文摘要
同一场景中的不同对象彼此或多或少相关，但是只有有限数量的这些关系值得注意。受到对象检测表现出色的DETR的启发，我们将场景图生成视为集合预测问题，并提出了具有编码器decoder架构的端到端场景图生成模型RELTR。关于视觉特征上下文的编码器原因是，解码器使用具有耦合主题和对象查询的不同类型的注意机制渗透了一组固定大小的三胞胎主题 - 主体对象。我们设计了一套预测损失，以执行端到端训练的地面真相与预测三胞胎之间的匹配。与大多数现有场景图生成方法相反，Reltr是一种单阶段方法，它仅使用视觉外观直接预测一组关系，而无需结合实体并标记所有可能的谓词。视觉基因组和开放图像V6数据集的广泛实验证明了我们模型的出色性能和快速推断。
Different objects in the same scene are more or less related to each other, but only a limited number of these relationships are noteworthy. Inspired by DETR, which excels in object detection, we view scene graph generation as a set prediction problem and propose an end-to-end scene graph generation model RelTR which has an encoder-decoder architecture. The encoder reasons about the visual feature context while the decoder infers a fixed-size set of triplets subject-predicate-object using different types of attention mechanisms with coupled subject and object queries. We design a set prediction loss performing the matching between the ground truth and predicted triplets for the end-to-end training. In contrast to most existing scene graph generation methods, RelTR is a one-stage method that predicts a set of relationships directly only using visual appearance without combining entities and labeling all possible predicates. Extensive experiments on the Visual Genome and Open Images V6 datasets demonstrate the superior performance and fast inference of our model.
