We present a deep neural network for removing undesirable shading features from an unconstrained portrait image, recovering the underlying texture. Our training scheme incorporates three regularization strategies: masked loss, to emphasize high-frequency shading features; soft-shadow loss, which improves sensitivity to subtle changes in lighting; and shading-offset estimation, to supervise separation of shading and texture. Our method demonstrates improved delighting quality and generalization when compared with the state-of-the-art. We further demonstrate how our delighting method can enhance the performance of light-sensitive computer vision tasks such as face relighting and semantic parsing, allowing them to handle extreme lighting conditions.
翻译:我们展示了一个深层的神经网络,从一个不受限制的肖像图像中去除不可取的阴影特征,恢复基本纹理。我们的培训计划包含三项正规化战略:掩盖损失,强调高频阴影特征;软阴影损失,这提高了对灯光微妙变化的敏感度;阴影估计,以监督阴影和纹理的分离。我们的方法表明,与最先进的相比,我们的方法提高了令人喜悦的质量和概括性。我们进一步展示了我们的喜悦方法如何能够提高对光敏感的计算机视觉任务(如面部光照亮和语义分解)的性能,使其能够处理极端的照明条件。