REL-SF4PASS：基于REL深度表示与球面融合的全景语义分割
REL-SF4PASS: Panoramic Semantic Segmentation with REL Depth Representation and Spherical Fusion
提出基于柱面坐标的REL深度表示和球面动态多模态融合（SMMF），为全景语义分割提供更有效的几何表征和融合策略，可迁移到其他全景感知任务。
原文依据 we propose REL-SF4PASS with our REL depth representation based on cylindrical coordinate and Spherical-dynamic Multi-Modal Fusion (SMMF)
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论全景语义分割（PASS）是计算机视觉中一个重要且具有挑战性的问题，旨在基于超广角视野提供完整的场景感知。大多数PASS方法通常关注RGB输入下的球面几何，或使用原始格式或HHA格式的深度信息，未能充分利用全景图像的几何特性。为解决这些不足，我们提出了REL-SF4PASS，其包含基于柱面坐标的REL深度表示和球面动态多模态融合（SMMF）。REL由校正深度、高程增益垂直倾角和横向方位角组成，以柱面坐标方式充分表示三维空间和表面法线方向。SMMF旨在确保不同全景图像区域融合的多样性，并减少ERP投影中圆柱侧面展开的断裂，使用不同的融合策略来匹配不同区域。实验结果表明，REL-SF4PASS在流行基准（如Stanford2D3D全景数据集）上显著提升了性能和鲁棒性。在所有3个折上平均交并比提升2.35%，面对三维扰动时性能方差降低约70%。
英文原文摘要 1254 字符
As an important and challenging problem in computer vision, Panoramic Semantic Segmentation (PASS) aims to provide complete scene perception based on an ultra-wide angle of view. Most PASS methods often focus on spherical geometry with RGB input or use the depth information in original or HHA format, which does not make full use of panoramic image geometry. To address these shortcomings, we propose REL-SF4PASS with our REL depth representation based on cylindrical coordinate and Spherical-dynamic Multi-Modal Fusion (SMMF). REL is made up of Rectified Depth, Elevation-Gained Vertical Inclination Angle, and Lateral Orientation Angle, which fully represent 3D space in cylindrical coordinate style and the surface normal direction. SMMF aims to ensure the diversity of fusion for different panoramic image regions and reduce the breakage of cylinder side surface expansion in ERP projection, using different fusion strategies to match different regions. Experimental results show that REL-SF4PASS considerably improves performance and robustness on popular benchmarks (e.g., Stanford2D3D Panoramic datasets). It gains 2.35% average mIoU improvement on all 3 folds and reduces the performance variance by approximately 70% when facing 3D disturbance.