多光渲染之光源分群的機器學習方法

Learning to cluster in rendering with many lights

王宇辰(Yu-Chen Wang)

指導教授 ： 莊永裕

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202101690

摘要

本篇論文提出了一個無偏差,在線的多光源渲染蒙地卡羅方法。我們的方法基於光源分層，我們採樣的概率基於我們的演算法已經得到的採樣結果。設計這樣的方法需要我們在嘈雜的採樣結果下做正確的分群決定，並能夠基於採樣結果得到好的採樣概率。我們的方法基於兩點精神。第一，我們從一個很小的光源分群出發找到一個較好的光源分群。這樣的方法能夠讓我們在很嘈雜的採樣結果下找到好的光源分群的方案。第二，我們採用隨機近似法讓我們從現有的採樣結果得到後驗概率來逼近目標分佈，證明出我們的方法一定能夠收斂。我們將我們的方法和現有的最先進的方法做對比，我們證明我們的方法在多光渲染的設定下能獲得更好的效果。

關鍵字

直接照明 ； 光線追蹤 ； 多光渲染 ； 強化學習

並列摘要

We present an unbiased online Monte Carlo method for rendering with many lights.Our method adapts both the hierarchical light clustering and the sampling distribution to our collected samples. Designing such a method requires us to make clustering decisions under noisy observation, and making sure that the sampling distribution adapts to our target. Our method is based on two key ideas: a coarse to ­fine clustering scheme that can find good clustering configurations even with noisy samples, and a discrete stochastic successive approximation method that starts from a prior distribution and provably converges to a target distribution. We compare to other state of ­the ­art light sampling methods, and show better results both numerically and visually.

並列關鍵字

Direct Illumination ； Ray Tracing ； Many­Light Rendering ； ReinforcementLearning

參考文獻

[1] B. Bitterli, C. Wyman, M. Pharr, P. Shirley, A. Lefohn, and W. Jarosz. Spatiotemporal reservoir resampling for realtime ray tracing with dynamic direct lighting. ACM Trans. Graph. (Proc. SIGGRAPH), 39(4):148–1, 2020.

[2] L. Bottou, F. E. Curtis, and J. Nocedal. Optimization methods for largescale machine learning. Siam Review, 60(2):223–311, 2018.

[3] C. Dachsbacher, J. Křivánek, M. Hašan, A. Arbree, B. Walter, and J. Novák. Scalable realistic rendering with manylight methods. Computer Graphics Forum, 33(1):88–104, 2014.

[4] C. Dachsbacher and M. Stamminger. Reflective shadow maps. In Proceedings of the 2005 symposium on Interactive 3D graphics and games, pages 203–231, 2005.

[5] K. Dahm and A. Keller. Learning light transport the reinforced way. In ACM SIGGRAPH 2017 Talks. Association for Computing Machinery, 2017.