Part-X-MLLM：部件感知的3D多模态大语言模型
Part-X-MLLM: Part-aware 3D Multimodal Large Language Model
AI总结：
本文提出部件感知的原生3D多模态大模型Part-X-MLLM，通过结构化可执行语法统一各类3D任务，经预训练与指令微调后，在多类3D任务上取得最优性能。
摘要：
我们提出Part-X-MLLM，这是一款原生3D多模态大语言模型，通过将各类3D任务表述为结构化可执行语法下的程序，实现了多样化3D任务的统一。给定RGB点云与自然语言提示，该模型以自回归方式生成连贯的单一token序列，编码部件级边界框、语义描述与编辑指令。这种结构化输出可作为通用接口，驱动下游几何感知模块完成基于部件的生成与编辑任务。通过将符号规划与几何合成解耦，我们的方法可通过单一语言原生前端控制任意兼容的几何引擎。我们预训练了双编码器架构以解耦结构与语义，并在大规模部件中心数据集上对模型进行指令微调。实验表明，该模型擅长生成高质量结构化规划，通过统一接口在定位问答、组合生成与局部编辑任务上取得了当前最优性能。项目主页：https://chunshi.wang/Part-X-MLLM/
英文摘要：
We introduce Part-X-MLLM, a native 3D multimodal large language model that unifies diverse 3D tasks by formulating them as programs in a structured, executable grammar. Given an RGB point cloud and a natural language prompt, our model autoregressively generates a single, coherent token sequence encoding part-level bounding boxes, semantic descriptions, and edit commands. This structured output serves as a versatile interface to drive downstream geometry-aware modules for part-based generation and editing. By decoupling the symbolic planning from the geometric synthesis, our approach allows any compatible geometry engine to be controlled through a single, language-native frontend. We pre-train a dual-encoder architecture to disentangle structure from semantics and instruction-tune the model on a large-scale, part-centric dataset. Experiments demonstrate that our model excels at producing high-quality, structured plans, enabling state-of-the-art performance in grounded Q\&A, compositional generation, and localized editing through one unified interface. Project page: https://chunshi.wang/Part-X-MLLM/
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
