arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.02259cs.CV
收藏

CTRL-C:带线分类的摄像机标定Transformer

CTRL-C: Camera calibration TRansformer with Line-Classification

Kakao Brain;Kookmin University(国民大学);POSTECH(浦项科技大学);KAIST(韩国科学技术院)

Jinwoo Lee, Hyunsung Go, Hyunjoon Lee, Sunghyun Cho, Minhyuk Sung, Junho Kim

摘要:

单图像摄像机标定是从一幅输入图像中估计摄像机参数的任务,如消失点、焦距和水平线。在这项工作中,我们提出了摄像机校准与线分类(CTRL-C),一个端到端的神经网络为基础的方法来单图像摄像机校准,直接估计摄像机参数从图像和一组线段。我们的网络采用Transformer架构,以端到端的方式捕获具有多模态输入的图像的全局结构。我们还提出了一个辅助任务的线分类训练网络提取的全局几何信息有效的线。我们的实验表明,CTRL-C优于以前的国家的最先进的方法在谷歌街景和SUN 360基准数据集。

英文摘要:

Single image camera calibration is the task of estimating the camera parameters from a single input image, such as the vanishing points, focal length, and horizon line. In this work, we propose Camera calibration TRansformer with Line-Classification (CTRL-C), an end-to-end neural network-based approach to single image camera calibration, which directly estimates the camera parameters from an image and a set of line segments. Our network adopts the transformer architecture to capture the global structure of an image with multi-modal inputs in an end-to-end manner. We also propose an auxiliary task of line classification to train the network to extract the global geometric information from lines effectively. Our experiments demonstrate that CTRL-C outperforms the previous state-of-the-art methods on the Google Street View and SUN360 benchmark datasets.

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。

↑