One stage relation detection (detects objects and relations simultaneously).

这个模型受启发于 [[CenterNet]] 和 [[OpenPose Using Part Affinity Fields]]，通过添加一个新的用于生成RAF的卷积头来获取物体之间的关系。

具体RAF是如何定义的，以及如何训练(Loss)：

Results

这篇看不懂。。。 Can effectively utilize structural information in the output space

Building an efficient structured representation that captures comprehensive semantic knowledge is a crucial step towards a deeper understanding of visual scenes

SGTR 是一种自上而下的方法，该方法首先使用基于Transformer的生成器来生成一组可学习的triplet queries (subject–predicate–object)，然后使用级联的triplet detector逐步完善这些查询并生成最终场景图。它还提出了一种基于结构化发生器的实体感知关系表示方法，该方法利用了关系的组成属性。

Top-down approach (SGTR):
- Starts with higher-level structures (triplet queries) and refines them
- Begins by generating complete subject-predicate-object triplet candidates
- Then progressively refines these triplets to match the image content
- Works with the complete structural units from the beginning
- Analogous to starting with a rough sketch of the entire tree and then refining each branch