解决3D实例分割中查询初始化与层深退化问题，提升定位与召回。
Beyond the Final Layer: Hierarchical Query Fusion Transformer with Agent-Interpolation Initialization for 3D Instance Segmentation
- 通过代理插值初始化生成兼顾位置与内容的鲁棒查询。
- 分层查询融合设计缓解深层网络中物体消失导致的召回下降。
- 在多个主流数据集上性能领先，适合3D场景理解任务。
3D实例分割旨在预测场景中一组物体实例，并以二值前景掩码和对应语义标签表示。当前基于Transformer的方法因结构简洁、无需手动设计几何特征且性能优越而备受关注。然而，现有方法在查询初始化阶段难以同时保持强位置与内容信息；此外，由于每层解码器均受监督，随着层数加深会出现物体消失现象。为此，本文提出面向3D实例分割的分层查询融合变压器（BFL），包含代理插值初始化模块，可生成兼具前景覆盖与内容学习能力的稳健查询；并设计分层查询融合解码器，保留低重叠查询，有效缓解深层网络中的召回率下降问题。在ScanNetV2、ScanNet200、ScanNet++和S3DIS等多个数据集上的大量实验表明，BFL显著优于现有方法。
原文摘要 · Abstract (English)
3D instance segmentation aims to predict a set of object instances in a scene and represent them as binary foreground masks with corresponding semantic labels. Currently, transformer-based methods are gaining increasing attention due to their elegant pipelines, reduced manual selection of geometric properties, and superior performance. However, transformer-based methods fail to simultaneously maintain strong position and content information during query initialization. Additionally, due to supervision at each decoder layer, there exists a phenomenon of object disappearance with the deepening of layers. To overcome these hurdles, we introduce Beyond the Final Layer: Hierarchical Query Fusion Transformer with Agent-Interpolation Initialization for 3D Instance Segmentation (BFL). Specifically, an Agent-Interpolation Initialization Module is designed to generate resilient queries capable of achieving a balance between foreground coverage and content learning. Additionally, a Hierarchical Query Fusion Decoder is designed to retain low overlap queries, mitigating the decrease in recall with the deepening of layers. Extensive experiments on ScanNetV2, ScanNet200, ScanNet++ and S3DIS datasets demonstrate the superior performance of BFL.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。