交互式意图驱动模型综合技术

75 浏览量更新于2024-08-26 收藏 2.18MB PDF 举报

"意图驱动的模型综合 - 研究论文" 本文主要介绍了一种创新的意图驱动的模型综合方法，该方法旨在通过交互式的直棱柱形构造空间来实现示例模型的结构和形状的动态调整。这种方法由南京大学新型软件技术国家重点实验室的研究团队提出，包括Mofei Song、Zhengxing Sun、Feiqian Zhang和Yan Zhang等人。在传统的模型综合过程中，设计者通常需要手动调整模型的各个参数以满足特定需求，这一过程既耗时又复杂。而意图驱动的模型综合方法则引入了一个全新的概念，即构造空间。这个构造空间是一种直棱柱形的几何框架，它允许用户通过基于草图的界面来设计和定义模型的全局尺寸和局部形状特征。用户可以绘制一个封闭曲线和一些骨架线来指定构造空间，这使得模型的设计更具灵活性和直观性。算法的核心是首先运用四边形细分算法，根据封闭曲线创建一个细分计划，使得模型的轮廓与用户绘制的曲线保持一致。接着，用户绘制的骨架线被用来指导模型内部的细节生成。这种方法的优势在于，它能够自动处理形状的复杂性和多样性，同时保留用户意图的清晰度。此外，文章还可能详细讨论了以下几点： 1. **模型变形机制**：如何通过改变构造空间的参数来实现模型的结构和形状变化，以及这种变化是如何平滑且连续地进行的。 2. **用户交互性**：用户界面的设计和使用，如何使得非专业用户也能有效地表达他们的设计意图。 3. **效率和精度**：算法在保证模型质量和设计精度的同时，如何提高合成速度和降低计算复杂性。 4. **应用领域**：可能探讨了该方法在建筑设计、产品设计、计算机图形学以及其他需要模型综合的领域的潜在应用。 5. **实验验证**：通过一系列的实验，验证了方法的有效性和实用性，比较了与其他传统模型综合方法的性能差异。 6. **未来工作**：可能讨论了对方法的进一步优化和扩展，以及可能遇到的挑战和未来的研究方向。这篇研究论文提供了一种新的、意图驱动的模型综合方法，通过构造空间的概念，使设计过程更加直观和高效，有望对模型设计和综合领域带来革新。

Intent-Driven Model Synthesis

Mofei Song

State Key Laboratory for Novel

Software Technology

Nanjing University

Nanjing, 210093, China

murphysong@gmail.com

Zhengxing Sun

∗

State Key Laboratory for Novel

Software Technology

Nanjing University

Nanjing, 210093, China

szx@nju.edu.cn

Feiqian Zhang

State Key Laboratory for Novel

Software Technology

Nanjing University

Nanjing, 210093, China

zhangfqjs@gmail.com

Yan Zhang

State Key Laboratory for Novel

Software Technology

Nanjing University

Nanjing, 210093, China

zhangyannju@nju.edu.cn

ABSTRACT

This paper presents an intent-driven model synthesis method.

The method introduces an interactive straight prismatic con-

struction space to realize the structure and shape variation

of the example model simultaneously. The construction s-

pace deﬁnes the global size and the local shape feature of

the desired model. Users can draw a closed curve and some

skeleton lines by a sketch-based interface to design the con-

struction space. Our algorithm ﬁrst uses a quadrangulation

algorithm to create a subdivision plane with the same con-

tour as the closed curve. And the drawn skeleton lines con-

trol the local orientation of split units in the plane. Then

it creates the construction space by sweeping the subdivi-

sion plane. Finally, it ﬁlls the construction space with the

deformed model pieces while maintaining the generalized ad-

jacent constraints, which are deﬁned according to the exam-

ple model. We demonstrate the eﬀectiveness of the approach

on large-scale complex models such as architecture, moun-

tains.

Categories and Subject Descriptors

I.3.5 [Computer Graphics]: Computational Geometry and

Object Modeling—Geometric algorithms, languages, and sys-

tems, Modeling packages

General Terms

Algorithms

Keywords

∗

Corresponding Author

Permission to make digital or hard copies of all or part of this work for

personal or classroom use is granted without fee provided that copies are not

made or distributed for proﬁt or commercial advantage and that copies bear

this notice and the full citation on the ﬁrst page. Copyrights for components

of this work owned by others than ACM must be honored. Abstracting with

credit is permitted. To copy otherwise, or republish, to post on servers or to

redistribute to lists, requires prior speciﬁc permission and/or a fee. Request

permissions from Permissions@acm.org.

VINCI ’13, August 17 - 18 2013, Tianjin, China

http://dx.doi.org/10.1145/2493102.2493107

construction space, structure variation, generalized adjacent

constraints

1. INTRODUCTION

Recently, large-scale complex models become an essential

way for oﬀering the richest user experience in the digital en-

tertainment or a virtual reality environment. Most 3D mod-

els are created by CAD modeling softwares such as Maya or

3D Max, however, learning how to use current geometric

modeling tools requires signiﬁcant training. And producing

large-scaled models remains a tedious work. Accordingly,

designing an intuitive and easy-used modeling tool is still a

challenging topic of computer graphics.

To simply the interaction, some sketch-based mo deling meth-

ods[8, 9, 18] have been presented. These methods focus on

creating coarse models, so they are not suitable for creating

the large-scale geometric structures. Compared with the

above approaches, the procedural modeling methods[17, 19]

are more powerful because they can turn a set of formal rules

to a complex geometry automatically. However, the mod-

eling process is often accompanied by the large amounts of

random parameters, which make it diﬃcult to control the

result. Actually, these automatic modeling methods do not

change the dominant position of the modelers. And they

should capture the users’ intent to drive the modeling pro-

cess and explore the required model in the potential solu-

tion set eﬀectively. As a result, how to control the proce-

dural modeling process becomes a central issue of computer

graphics.

To solve this problem, some interactive procedural modeling

methods have been proposed and users can express their in-

tent interactively to guide the modeling process. Constraint-

based procedural modeling methods[15, 21] use an optimiza-

tion approach to generate the required result, which satis-

ﬁes some global constraints given by users. These methods

provide a huge solution space for the enormous structure

variation, but due to lack of freeform deformation, users

cannot control the local shape and the diversity limits on

existing or predeﬁning structure. Bokeloh et al[2, 3, 4, 16]

combine the symmetrical replication rule and the elastic de-

下载后可阅读完整内容，剩余9页未读，立即下载

weixin_38530202

粉丝: 2
资源: 876

交互式意图驱动模型综合技术

电动汽车整车控制策略模型

Domain-Driven Design领域驱动设计

文字人工智能交互模型及代码

战略驱动型并购与研究.ppt

基于神经网络的驾驶员制动意图辨识方法研究.pdf

卡扣端子分选机全套图纸-几何模型

灰关联贴近度决策模型：功能驱动与差异驱动结合

特征模型与Web服务驱动的工程知识库共享机制

用户点击模型驱动的查询扩展技术在信息检索中的应用

探索Microsoft Copilot：GPT-4驱动的多语言对话AI模型

最新资源