面向手语生成的语义一致引导焦点-全局扩散模型
Focal-General Diffusion Model with Semantic Consistent Guidance for Sign Language Production
两阶段去噪框架把局部关节级建模（自适应图卷积）与全局序列精修（Transformer）解耦，并用语义一致引导把文本语义监督注入扩散训练，这一「局部-全局分层去噪 + 语义引导」思路可迁移到人体姿态/动作序列生成等结构化输出任务。
原文依据 a pioneering two-stage denoising framework that harmonizes local joint-level dependencies and global coherence
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论手语生成（SLP）旨在将口语翻译为手语序列，其主要挑战在于从离散的 gloss 生成连贯自然的手语姿态（G2P）。现有 G2P 方法通常将每个姿态视为不可分割的整体，限制了捕捉细粒度关节级依赖的能力，从而降低了姿态质量。为此，我们提出焦点-全局扩散模型（FGDM），其特点是开创性的两阶段去噪框架，协调局部关节级依赖与全局连贯性。具体而言，在焦点阶段，一种新颖的自适应手语图卷积网络（ASGCN）基于上下文关联、骨架拓扑与语义条件对每个姿态进行自适应建模，确保局部细节的精确生成。在全局阶段，基于 Transformer 的模块对整个姿态序列进行精修，以增强全局连贯性与自然度。此外，我们引入语义一致引导（SCG）机制，将语义监督无缝融入扩散训练，促使生成的姿态序列与其目标 gloss 语义之间实现更紧密的对齐。在 PHOENIX14T 与 USTC-CSL 上的大量实验表明，FGDM 达到了 SOTA 性能。项目主页见 https://yuyiheng-eu.github.io/fgdm/。
英文原文摘要 1315 字符
Sign Language Production (SLP) aims to translate spoken language into sign sequences, where the main challenge lies in generating coherent and natural poses from discrete glosses (G2P). Existing G2P methods typically treat each pose as an indivisible unit, limiting their ability to capture fine-grained joint-level dependencies and thus degrading pose quality. To address this, we propose the Focal-General Diffusion Model (FGDM), characterized by a pioneering two-stage denoising framework that harmonizes local joint-level dependencies and global coherence. Specifically, in the Focal stage, a novel Adaptive Sign GCN (ASGCN) adaptively models each pose based on contextual correlations, skeletal topology, and semantic conditions, ensuring precise generation of local details. In the General stage, a Transformer-based module refines the entire pose sequence to enhance global coherence and naturalness. Moreover, we introduce a Semantic Consistent Guidance (SCG) mechanism that seamlessly integrates semantic supervision into diffusion training, enforcing tighter alignment between generated pose sequences and their intended gloss semantics. Extensive experiments on PHOENIX14T and USTC-CSL demonstrate that FGDM achieves SOTA performance. Our project page is available at https://yuyiheng-eu.github.io/fgdm/.