用于蒙特卡洛去噪的鲁棒平均网络
Robust Average Networks for Monte Carlo Denoising
AI总结:
该研究提出通过插入鲁棒平均块、修改架构与损失函数并保持训练时时间连接活跃,将空间去噪神经网络转为时空型,实现提升图像质量、减少时间闪烁的效果。
摘要:
我们提出一种通过修改网络架构和损失函数,将去噪神经网络从空间型转换为时空型的方法。我们在网络图的任意深度插入鲁棒平均块,每个块利用可训练权重执行潜在空间插值,并对来自网络前序空间组件的图像表示序列进行处理。训练期间,通过迫使网络从输入序列的子集预测去噪帧,保持时间连接活跃。利用时间一致性去噪可提升图像质量,且无论场景或图像复杂度如何,均能减少时间闪烁。
英文摘要:
We present a method for converting denoising neural networks from spatial into spatio-temporal ones by modifying the network architecture and loss function. We insert Robust Average blocks at arbitrary depths in the network graph. Each block performs latent space interpolation with trainable weights and works on the sequence of image representations from the preceding spatial components of the network. The temporal connections are kept live during training by forcing the network to predict a denoised frame from subsets of the input sequence. Using temporal coherence for denoising improves image quality and reduces temporal flickering independent of scene or image complexity.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。