Enric Moreu, Kevin McGuinness, Diego Ortego, Noel E. O'Connor
TL;DR本文提出一种使用基于合成数据集的域随机化方法进行物体计数的方法,避免了昂贵的 3D 艺术家手动生成照片般逼真的数据集,增加了数据集的多样性和泛化性,实现了较好的性能表现。
Abstract
Recently, the use of synthetic datasets based on game engines has been shown
to improve the performance of several tasks in computer vision. However, these
datasets are typically only appropriate for the specific