SAQN：面向三维指代表达分割的语义自适应查询网络
SAQN: Semantic-based Adaptive Query Network for 3D Referring Expression Segmentation
用每个语义类一个可学习查询向量替代从点云生成查询，把查询数量从数万级降到类别级，并用隐式类别对齐绕开匈牙利匹配；做 3D 指代分割或点云开放词表分割时可直接复用这套查询构造与补充查询融合策略。
原文依据 SAQN employs a learnable query vector for each semantic class.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论三维指代表达分割（3D-RES）旨在根据语言描述分割点云中的物体。与二维中普遍采用可学习查询嵌入的做法不同，近期的 3D-RES 方法通常直接从三维点生成查询。然而，这种将查询与原始点云直接耦合的做法带来了新的挑战：从海量点云数据中导出的查询数量大到不切实际，并且依赖非确定性的采样算法。本文提出语义自适应查询网络（SAQN），为 3D-RES 引入一种新的查询策略。SAQN 不再从点生成查询，而是为每个语义类别使用一个可学习查询向量。该做法大幅减少了查询数量，同时通过隐式类别对齐保留了避免匈牙利匹配的优势。此外，为应对语义类别内部潜在的跨物体歧义，我们引入补充查询，使其与每个类别查询自适应融合，以消解歧义并丰富表征。大量实验表明，SAQN 在减少查询数量的同时达到了最优性能。
英文原文摘要 1202 字符
3D Referring Expression Segmentation (3D-RES) aims to segment objects in point clouds according to language descriptions. Unlike common practices in 2D that utilize learnable query embeddings, recent 3D-RES methods typically generate queries directly from 3D points. However, this direct coupling of queries to raw point clouds introduces new challenges: an impractically large number of queries derived from massive point cloud data and a reliance on non-deterministic sampling algorithms. In this paper, we propose a Semantic-based Adaptive Query Network (SAQN), which introduces a novel query strategy for 3D-RES. Instead of generating queries from points, SAQN employs a learnable query vector for each semantic class. This approach drastically reduces the number of queries while maintaining the advantage of avoiding Hungarian matching through implicit class alignment. Additionally, to address potential cross-object ambiguity within semantic classes, we introduce supplementary queries that are adaptively fused with each class query to disambiguate and enrich representations. Comprehensive experiments show that SAQN achieves state-of-the-art performance while reducing the number of queries.