VALSE 论文速览 第174期：PlaneRecTR: Query Learning for Plane Recovery

论文题目：

PlaneRecTR: Unified Query Learning for 3D Plane Recovery from a Single View

作者列表：

史静佳 (国防科技大学)、智帅峰 (国防科技大学)、徐凯 (国防科技大学)

B站观看网址：

https://www.bilibili.com/video/BV1im421n7KK/

论文摘要：

3D plane recovery from a single image can usually be divided into several subtasks of plane detection, segmentation, parameter estimation and possibly depth estimation. Previous works tend to solve it by either extending the RCNN-based segmentation network or the dense pixel embedding-based clustering framework. However, none of them tried to integrate above related subtasks into a unified framework but treated them separately and sequentially, which we suspect is potentially a main source of performance limitation for existing approaches. Motivated by this finding and the success of query-based learning in enriching reasoning among semantic entities, in this paper, we propose PlaneRecTR, a Transformer-based architecture, which for the first time unifies all subtasks related to single view plane recovery with a single compact  model. Extensive quantitative and qualitative experiments demonstrate that our proposed unified learning achieves mutual benefits across subtasks, obtaining a new state-of-the-art performance on public ScanNet and NYUv2-Plane datasets. 

参考文献：

[1] Jingjia Shi, Shuaifeng Zhi, and Kai Xu. "PlaneRecTR: Unified Query Learning for 3D Plane Recovery from a Single View." Proceedings of the IEEE/CVF International Conference on Computer Vision. 2023.

论文链接：

[https://openaccess.thecvf.com/content/ICCV2023/html/Shi_PlaneRecTR_Unified_Query_Learning_for_3D_Plane_Recovery_from_a_ICCV_2023_paper.html]

代码链接：

[https://github.com/SJingjia/PlaneRecTR]

视频讲者简介：

史静佳，国防科技大学计算机学院直博研究生，本科毕业于中南大学智能科学与技术专业，目前的研究方向为结构化场景重建。

特别鸣谢本次论文速览主要组织者：

月度轮值AC：胡迪 (中国人民大学)