Browse > Article
http://dx.doi.org/10.7583/JKGS.2020.20.3.15

Reinforcement Learning-based Approach for Lego Puzzle Generation  

Park, Cheolseong (Dept. of Computer Science, Sangmyung Univ.)
Yang, Heekyung (Div. of SW Convergence, Sangmyung Univ.)
Min, Kyungha (Dept. of Computer Science, Sangmyung Univ.)
Publication Information
Journal of Korea Game Society / v.20, no.3, 2020 , pp. 15-24 More about this Journal
Abstract
We present a reinforcement learning-based framework for generating 2D Lego puzzle from input pixel art images. We devise heuristics for a proper Lego puzzle as stability and efficiency. We also design a DQN structure and train it to maximize the heuristics of 2D Lego puzzle. In legorization stage, we complete the layout of Lego puzzle by adding a Lego brick to the input image using the trained DQN. During this process, we devise a region of interest to reduce the computational loads of the legorization. Using this approach, our framework can present a very high resolutional Lego puzzle.
Keywords
Legorization; Reinforcement learning; Heuristic;
Citations & Related Records
연도 인용수 순위
  • Reference
1 R. Gower, A. Heydtmann, and H. Petersen, "Lego: Automated model construction," in Proc. of 32nd European Study Group with Industry 1998, 1998, pp. 81-94.
2 P. Petrovic, "Solving lego brick layout problem using evolutionary algorithms," in Proc. of Norwegian Conference on Computer Science, 2001, pp. 87-97.
3 L. van Zijl and E. Smal, "Cellular automata with cell clustering," in Proc. of Automata 2008, 2008, pp. 425-440.
4 R. Testuz, Y. Schwartzburg, and M. Pauly, "Automatic generation of constructable brick sculptures," in Proc. of Eurographics (Short Papers) 2013, 2013, pp. 81-84.
5 S. Ono, A. Andre, and Y. Chang, "Automatic generation of lego from the polygonal data," in Proc. of International workshop on advanced image technology 2013, 2013, pp. 262-267.
6 S. Lee, J. Kim, J. W. Kim, and B. R. Moon, "Finding an optimal lego brick layout of voxelized 3d object using a genetic algorithm," in Proc. of the 2015 Annual Conference on Genetic and Evolutionary Computation, 2015, pp. 1215-1222.
7 M. Zhang, Y. Igarashi, Y. Kanamori, and J. Mitani, "Designing miniblock artwork from colored mesh," in Proc. of Smart Graphics 2015, 2015, p. 2.
8 G. Yun, C. Park, H. Yang, and K. Min, "Legorization with multiheight bricks from silhouette-fitted voxelization," in Proc. of Computer Graphics International 1986 (Short paper), 2017, p. 40.
9 S. J. Luo, Y. Yue, C. K. Huang, Y. H. Chung, S. Imai, T. Nishita, and B. Y. Chen, "Legolization: Optimizing lego designs," ACM Transactions on Graphics, vol. 34, no. 6, p. 222, 2015.
10 T. Kosaki, H. Tedenuma, and T. Maekawa, "Automatic generation of lego building instructions from multiple photographic images of real objects," Computer-Aided Design, vol. 70, no. C, pp. 13-22, 2016.   DOI
11 M. Zhang, J. Mitani, Y. Kanamori, and Y. Fukui, "Blocklizer: Interactive design of stable mini block artwork," in Proc. of ACM SIGGRAPH 2014 Posters, 2014, p. 18.
12 M. Kuo, Y. Lin, H. Chu, R. Lee, and Y. Yang, "Pixel2brick: Constructing brick sculptures from pixel art." Computer Graphics Forum, vol. 34, no. 7, pp. 339-348, 2015.   DOI
13 Reinforcement Learning, https://en.wikipedia.org/wiki/Reinforcement_learning.
14 T. Schaul, J. Quan, I. Antonoglou, and D. Silver, "Prioritized experience replay," Proc. of International Conference on Learning Representation (ICLR) 2016.
15 V. Mnih, K. Kavukcuoglu, D. Silver, A. Rusu, J. Veness, M. Bellemare, A. Graves, M. Riedmiller, A. Fidjeland, G. Ostrovski, S. Petersen, C. Beattie, A. Sadik, I. Antonoglou, H. King, D. Kumaran, D. Wierstra, S. Legg, and D. Hassabis, "Human-level control through deep reinforcement learning,"Nature, Vol. 518, No. 7540, pp. 529-533, 2015.   DOI
16 H. Hasselt, A. Guez and D. Silver, "Deep reinforcement learning with double Q-learning," Proc. of AAAI conference on Artificial Intelligence 2016, pp. 2094-2100, 2016.
17 Bellemare, M., Dabney, W., and Munos, R., "A distributional perspective on reinforcement learning," Proc. of International Conference on Machine Learning 2017, pp. 449-458, 2017.