Implementation of Autoencoder based Neural Network for Realtime Ray Tracing
AUTHORS
HakHyun Lee,Dept. Computer Engineering, SeoKyeong Univ., 124, Seogyeong-ro, Seongbuk-gu, Seoul, Republic of Korea
Chelwon Jo,Dept. Computer Engineering, SeoKyeong Univ., 124, Seogyeong-ro, Seongbuk-gu, Seoul, Republic of Korea
Kwang-Yeob Lee,Dept. Computer Engineering, SeoKyeong Univ., 124, Seogyeong-ro, Seongbuk-gu, Seoul, Republic of Korea
ABSTRACT
This paper proposes a denoising neural network for real-time ray tracing. The ray tracing method is applied in graphics to increase reality and in particular, Monte Carlo Rendering is most effective. However, ray tracing that applies Monte Carlo Rendering has a steep rise in the amount of calculations with the increase of the number of rays. Therefore, in order to solve this problem, various methods are being proposed to reduce the number of rays and to decrease the occurring noise. In this paper, an autoencoder-based neural network that can effectively remove noise while using a small number of rays was implemented. An autoencoder that uses a 1×1 convolution in creating the last feature map was proposed to significantly lower the amount of calculation. The proposed structure can handle 8196 spp ray-tracing image in 20 seconds at 64 spp.
KEYWORDS
Ray tracing, Denoising, Monte carlo rendering, Neural network, Autoencoder
REFERENCES
[1] VEACH, Eric, GUIBAS, and Leonidas J., “Optimally combining sampling techniques for Monte Carlo rendering,” Proceedings of the 22nd annual conference on Computer graphics and interactive techniques, ACM, pp.419-428, (1995) DOI: 10.1145/218380.218498(CrossRef)(Google Scholar)
[2] Wang Z., Bovik A. C., Sheikh H. R., and Simoncelli E. P., “Image quality assessment: from error visibility to structural similarity,” IEEE transactions on image processing, vol.13, no.4, pp.600-612, (2004) DOI: 10.1109/TIP.2003.819861(CrossRef)(Google Scholar)
[3] Hore A. and Ziou D., “Image quality metrics: PSNR vs. SSIM,” 2010 20th International Conference on Pattern Recognition, IEEE, pp.2366-2369, (2010) DOI: 10.1109/ICPR.2010.579(CrossRef)(Google Scholar)
[4] Vincent P., Larochelle H., Lajoie I., Bengio Y., and Manzagol P. A., “Stacked denoising autoencoders: Learning useful representations in a deep network with a local denoising criterion,” Journal of machine learning research, vol.11, pp.3371-3408, (2010) DOI: 10.1016/j.mechatronics.2010.09.004(CrossRef)(Google Scholar)
[5] Veach E., and Guibas L. J., “Optimally combining sampling techniques for Monte Carlo rendering,” Proceedings of the 22nd annual conference on Computer graphics and interactive techniques, pp.419-428, (1995) DOI: 10.1145/218380.218498(CrossRef)(Google Scholar)
[6] Nair V., and Hinton G. E., “Rectified linear units improve restricted Boltzmann machines,” Proceedings of the 27th international conference on machine learning (ICML-10), pp.807-814, (2010)
[7] Mao,X. J., Shen C., and Yang Y. B., “Image restoration using convolutional auto-encoders with symmetric skip connections,” arXiv preprint arXiv:1606.08921, (2016)
[8] Maas A. L., Hannun A. Y., and Ng A. Y., “Rectifier nonlinearities improve neural network acoustic models,” Proc. Icml, vol.30, no.1, p.3, (2013)
CITATION
COPYRIGHT
© 2020 HakHyun Lee et al. Published by Global Vision Press. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CCBY4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.