| 摘要: |
| 目的 当前人工智能生成内容技术(AIGC)已步入全方位商业化阶段，AIGC正在成为智能设计的核心支持手段。为提高设计效率和设计质量，顺应人工智能时代不断变化的复杂设计需求，对AIGC支持下的智能设计方法进行综述。方法 通过文献研究，对AIGC技术的基本原理和主要特点进行深入分析，归纳出基于数据挖掘与分析、知识推理与学习、算法演化与模拟的智能设计方法，并得出数据预处理、语境感知调谐、跨模态交互设计、模型构建与评估、动态化学习与反馈优化等AIGC支持下智能设计的关键步骤。结果 AIGC在工业设计流程迭代、传媒创作人机深度融合和娱乐交互方式研发等领域对智能设计方法的变革具有重大意义。结论 对AIGC技术支持下的智能设计方法进行展望，得出虚实结合下的智能化数字孪生设计、全球协同下的智能设计社会变革、隐私安全需求下的交互反馈模式设计3个方面的趋势。 |
| 关键词: 智能设计方法 人工智能生成内容 应用领域 未来展望 综述 |
| DOI：10.19554/j.cnki.1001-3563.2023.24.003 |
| 分类号:TB472 |
| 基金项目:国家自然科学基金面上项目(52275234);教育部人文社科基金(22YJA760055) |
|
| State of Arts and Development of Intelligent Design Methods under the AIGC Trend |
|
LU Zhao-lin1,2, SONG Xin-heng1,2, JIN Yu-cheng3
|
|
(1. School of Design and Art, Beijing Institute of Technology, Beijing 100081, China;2. Industrial Manufacturing Art Innovation Design Key Laboratory of Ministry of Industry and Information Technology, Beijing 100081, China;3. Department of Computer Science, Hong Kong Baptist University, Hong Kong 999077, China)
|
| Abstract: |
| Artificial Intelligence Generated Content (AIGC) technology has stepped into an all-round commercialization stage. AIGC is becoming a core support measure for intelligent design. The work aims to review the intelligent design methods supported by AIGC to improve the design efficiency and design quality, and to comply with the ever-changing complex design needs in the age of artificial intelligence. Through literature research, basic principles and main features of AIGC were analyzed deeply. Intelligent design methods based on data mining and analysis, knowledge reasoning and learning, algorithm evolution and simulation were summarized. Key steps of AIGC-supported intelligent design:data preprocessing, context-aware tuning, cross-modal interaction design, model construction and evaluation, dynamic learning and feedback optimization were derived. AIGC was of great significance for the transformation of intelligent design methods in industrial design process iteration, deep integration of man and machine in media creation entertainment interaction development. Finally, the intelligent design method supported by AIGC is prospected, and three trends are drawn:intelligent digital twin design under the combination of virtual and real, the social change of intelligent design under the global collaboration, the design of interaction and feedback mode under the demand of privacy and security. |
| Key words: Syntax Networks for Code Generation and Semantic Parsing[C]// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1:Long Papers). Vancouver, Canada. Stroudsburg, PA, USA:Association for Computational Linguistics, 2017:1139-1149. |
渝公网安备 50010702501716号