模仿学习:一种新人工生命动画方法

班晓娟; 徐卓然; 刘浩

doi:10.3724/SP.J.1004.2012.00518

模仿学习:一种新人工生命动画方法

doi: 10.3724/SP.J.1004.2012.00518

1.
北京科技大学计算机与通信工程学院北京 100083

详细信息

通讯作者:
徐卓然北京科技大学计算机科学与技术系硕士研究生. 主要研究方向为人工智能和神经网络. E-mail: xuzhuoran0106@gmail.com

计量
- 文章访问数: 2232
- HTML全文浏览量: 67
- PDF下载量: 1284
- 被引次数: 0
出版历程
- 收稿日期: 2011-07-18
- 修回日期: 2011-11-10
- 刊出日期: 2012-04-20

Imitation Learning: A New Approach in Artificial Life Animation

1.
School of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing 100083

摘要

摘要: 提出一种新的人工生命动画方法—模仿学习. 模仿是一种非常有效的掌握运动技能的学习方式. 一项运动技能为无数个相关运动序列的集合. 通过模仿代表性运动序列,将蕴含的局部运动技能泛化,可获得完整的运动技能. 模仿学习以运动相似度匹配和简单--复杂行为方法论为核心,并以进化计算为优化方法. 模仿学习降低进化计算对传统评价函数的依赖,减少评价函数设计时间,提高优化复杂目标的能力,因此提高了制作效率. 基于PhysX仿真平台,本文以人工猫的着陆行为验证了本文方法的有效性,并取得了良好的效果.
- 模仿学习 /
- 简单--复杂行为方法论 /
- 运动相似度匹配 /
- 进化计算 /
- 人工生命动画
Abstract: This paper proposes a new artificial life animation approach—imitation learning. Imitation is a highly effective learning method for acquiring motion skill which can be regarded as a set of numerous motion sequences. Imitating representative motion sequences to acquire the local motion skill and generalizing them can achieve the entire motion skill. The cores of imitation learning are motion similarity and simple-compose behavior methodology, and evolutionary computation is used as an optimization method. Imitation learning decreases the dependence of evolutionary computation on traditional fitness function and the time spent on designing a suitable fitness function. It also increases the ability of optimizing complex goal. So it increases the efficiency of producing animation. We verify our method by training an artificial cat robot to learn landing behavior based on PhysX simulation framework, which achieves a good result.
- Imitation learning /
- simple-compose behavior /
- motion similarity matching /
- evolutionary computation /
- artificial life animation

HTML全文

参考文献(1)

[1]

Tu X, Terzopoulos D. Artificial fishes: physics, locomotion, perception, behavior. In: Proceedings of the 21st Annual Conference on Computer Graphics and Interactive Techniques. Orlando, USA: ACM, 1994. 43-50[2] Wu J C, Popovic Z. Realistic modeling of bird flight animations. ACM Transactions on Graphics, 2003, 22(3): 888-895[3] Liu C K, Hertzmann A, Popovic Z. Learning physics-based motion style with nonlinear inverse optimization. ACM Transactions on Graphics, 2005, 24(3): 1071-1081[4] Sims K. Evolving virtual creatures. In: Proceedings of the 21st Annual Conference on Computer Graphics and Interactive Techniques. Orlando, USA: ACM, 1994. 15-22[5] Sims K. Evolving 3D morphology and behavior by competition. Artificial Life, 1994, 1(4): 353-372[6] Tanev I, Ray T, Buller A. Automated evolutionary design, robustness, and adaptation of sidewinding locomotion of a simulated snake-like robot. IEEE Transactions on Robotics, 2005, 21(4): 632-645[7] Secretan J, Beato N, D'Ambrosio D B, Rodriguez A, Campbell A, Folsom-Kovarik J T, Stanley K O. Picbreeder: a case study in collaborative evolutionary exploration of design space. Evolutionary Computation, 2011, 19(3): 373-403[8] Marriott C, Parker J, Denzinger J. Imitation as a mechanism of cultural transmission. Artificial Life, 2010, 16(1): 21-37[9] Pullen K, Bregler C. Motion capture assisted animation: texturing and synthesis. ACM Transactions on Graphics, 2002, 21(3): 501-508[10] Liu F, Zhuang Y, Wu F, Pan Y. 3D motion retrieval with motion index tree. Computer Vision and Image Understanding, 2003, 92(2-3): 265-284[11] Lee J, Chai J, Reitsma P S A, Hodgins J, Pollard N S. Interactive control of avatars animated with human motion data. ACM Transactions on Graphics, 2002, 21(3): 491-500[12] Jolliffe I T. Principle Component Analysis. New York: Spring, 1986[13] Guttman A. R-trees: a dynamic index structure for spatial searching. ACM SIGMOD Record, 1984, 14(2): 47-57[14] Mukai N, Watanabe T, Jun F. R-tree based optimization algorithm for dynamic transport problem. Lecture Notes in Computer Science. 2006, 4252: 1095-1102[15] Li C, Pradhan G, Zheng S, Prabhakaran B. Indexing of variable length multi-attribute motion data. In: Proceedings of the 2nd ACM International Workshop on Multimedia Databases. Washington D. C., USA: ACM, 2004. 75-84[16] Kahveci T, Singh A, Gurel A. Similarity searching for multi-attribute sequences. In: Proceedings of the 14th International Conference on Scientific and Statistical Database Management. Edinburgh, UK: IEEE, 2002. 175-184[17] Lee S L, Chun S J, Kim D H, Lee J H, Chung C W. Similarity search for multidimensional data sequences. In: Proceedings of the 16th International Conference on Data Engineering. San Diego, USA: IEEE, 2000. 599-608[18] Gomez F, Schmidhuber J, Miikkulainen R. Accelerated neural evolution through cooperatively coevolved synapses. Journal of Machine Learning Research, 2009, 9: 937-965[19] Stanley K O, Miikkulainen R. Evolving neural networks through augmenting topologies. Evolutionary Computation, 2002, 10(2): 99-127[20] Dürr P, Mattiussi C, Floreano D. Neuroevolution with analog genetic encoding. Lecture Notes in Computer Science, 2006, 4193: 671-680

施引文献

资源附件(0)

访问统计