arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2108.07533cs.CV
收藏

研究多边形形状分解为点集的Transformers

Investigating transformers in the decomposition of polygonal shapes as point collections

Delft University of Technology(代尔夫特理工大学)

Andrea Alfieri, Yancong Lin, Jan C. van Gemert

摘要:

Transformers可以通过两种方法生成预测:1.通过以先前的序列元素为条件来自回归地调节每个序列元素,或者2.直接并行产生输出序列。虽然研究主要探讨了NLP中顺序任务的这种差异,但我们研究了视觉集预测任务中自回归和并行预测之间的差异,特别是图像中的多边形形状,因为多边形代表了许多类型的对象,例如建筑物或飞行器的障碍物。这对于深度学习架构来说是具有挑战性的,因为多边形可以由不同的点组成。我们提供的证据自然秩序的重要性,Transformers,并显示出的好处,分解复杂的多边形的点的集合中的自回归的方式。

英文摘要:

Transformers can generate predictions in two approaches: 1. auto-regressively by conditioning each sequence element on the previous ones, or 2. directly produce an output sequences in parallel. While research has mostly explored upon this difference on sequential tasks in NLP, we study the difference between auto-regressive and parallel prediction on visual set prediction tasks, and in particular on polygonal shapes in images because polygons are representative of numerous types of objects, such as buildings or obstacles for aerial vehicles. This is challenging for deep learning architectures as a polygon can consist of a varying carnality of points. We provide evidence on the importance of natural orders for Transformers, and show the benefit of decomposing complex polygons into collections of points in an auto-regressive manner.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑