Label Denoising Adversarial Network (LDAN) for Inverse Lighting of Face Images
This addresses lighting estimation for applications like image editing and forgery detection, but it is incremental as it builds on existing GAN and CNN techniques.
The authors tackled the problem of lighting estimation from single face images by proposing a Label Denoising Adversarial Network (LDAN) that uses synthetic data with accurate ground truth to train a CNN, resulting in more consistent lighting parameters and being 100,000 times faster than prior methods.
Lighting estimation from face images is an important task and has applications in many areas such as image editing, intrinsic image decomposition, and image forgery detection. We propose to train a deep Convolutional Neural Network (CNN) to regress lighting parameters from a single face image. Lacking massive ground truth lighting labels for face images in the wild, we use an existing method to estimate lighting parameters, which are treated as ground truth with unknown noises. To alleviate the effect of such noises, we utilize the idea of Generative Adversarial Networks (GAN) and propose a Label Denoising Adversarial Network (LDAN) to make use of synthetic data with accurate ground truth to help train a deep CNN for lighting regression on real face images. Experiments show that our network outperforms existing methods in producing consistent lighting parameters of different faces under similar lighting conditions. Moreover, our method is 100,000 times faster in execution time than prior optimization-based lighting estimation approaches.