3D-GPT: Procedural 3D MODELING WITH LARGE LANGUAGE MODELS 3D-GPT： 使用大型語言模型進(jìn)行程序化 3D 建模
2023.10
論文主頁
論文地址
代碼地址（未放）
一句話生成3D世界，未公布代碼已獲141星！或引發(fā)3D建模行業(yè)革命
Abstract
In the pursuit of efficient automated content creation, procedural generation, leveraging modifiable parameters and rule-based systems, emerges as a promising approach. Nonetheless, it could be a demanding endeavor, given its intricate nature necessitating a deep understanding of rules, algorithms, and parameters. To reduce workload, we introduce 3D-GPT, a framework utilizing large language models~(LLMs) for instruction-driven 3D modeling. 3D-GPT positions LLMs as proficient problem solvers, dissecting the procedural 3D modeling tasks into accessible segments and appointing the apt agent for each task. 3D-GPT integrates three core agents: the task dispatch agent, the conceptualization agent, and the modeling agent. They collaboratively achieve two objectives. First, it enhances concise initial scene descriptions, evolving them into detailed forms while dynamically adapting the text based on subsequent instructions. Second, it integrates procedural generation, extracting parameter values from enriched text to effortlessly interface with 3D software for asset creation. Our empirical investigations confirm that 3D-GPT not only interprets and executes instructions, delivering reliable results but also collaborates effectively with human designers. Furthermore, it seamlessly integrates with Blender, unlocking expanded manipulation possibilities. Our work highlights the potential of LLMs in 3D modeling, offering a basic framework for future advancements in scene generation and animation.
在追求高效的自動(dòng)內(nèi)容創(chuàng)建過程中，利用可修改參數(shù)和基于規(guī)則的系統(tǒng)進(jìn)行程序生成是一種很有前途的方法。
然而，由于其復(fù)雜性，需要對(duì)規(guī)則、算法和參數(shù)有深入的了解，這可能是一項(xiàng)艱巨的工作。
為了減少工作量，我們引入了 3D-GPT 框架，該框架利用大型語言模型（LLM）進(jìn)行指令驅(qū)動(dòng)的 3D 建模。
3D-GPT 將大型語言模型定位為熟練的問題解決者，將程序化三維建模任務(wù)分解為可訪問的片段，并為每個(gè)任務(wù)指定合適的agent。
3D-GPT 集成了三個(gè)核心agent：
- 任務(wù)派遣agent;
- 概念化agent;
- 建模agent。
它們共同實(shí)現(xiàn)了兩個(gè)目標(biāo)：
- 首先，它增強(qiáng)了簡(jiǎn)潔的初始場(chǎng)景描述，將其發(fā)展為詳細(xì)的形式，同時(shí)根據(jù)后續(xù)指令動(dòng)態(tài)調(diào)整文本。
- 其次，它整合了程序生成功能，從豐富的文本中提取參數(shù)值，從而輕松地與三維軟件對(duì)接，進(jìn)行資產(chǎn)創(chuàng)建。
我們的實(shí)證調(diào)查證實(shí)，3D-GPT 不僅能解釋和執(zhí)行指令，提供可靠的結(jié)果，還能與人類設(shè)計(jì)師有效協(xié)作。此外，它還能與 Blender 無縫集成，從而實(shí)現(xiàn)更多的操作可能性。我們的工作彰顯了 LLM 在三維建模中的潛力，為未來場(chǎng)景生成和動(dòng)畫制作的進(jìn)步提供了一個(gè)基本框架。文章來源：http://www.zghlxwxcb.cn/news/detail-843233.html
簡(jiǎn)評(píng)
概念很有意思，但是實(shí)際很簡(jiǎn)單，本質(zhì)為blender + python，由gpt生成python代碼。加上項(xiàng)目未發(fā)布開源代碼，噱頭略大于實(shí)際。文章來源地址http://www.zghlxwxcb.cn/news/detail-843233.html
到了這里，關(guān)于【LLM+三維場(chǎng)景】3D-GPT: Procedural 3D MODELING WITH LARGE LANGUAGE MODELS的文章就介紹完了。如果您還想了解更多內(nèi)容，請(qǐng)?jiān)谟疑辖撬阉鱐OY模板網(wǎng)以前的文章或繼續(xù)瀏覽下面的相關(guān)文章，希望大家以后多多支持TOY模板網(wǎng)！
