Depth map prediction from a single image with generative adversarial nets

A depth map is a fundamental component of 3D construction. Depth map prediction from a single image is a challenging task in computer vision. In this paper, we consider the depth prediction as an image-to-image task and propose an adversarial convolutional architecture called the Depth Generative Adversarial Network (DepthGAN) for depth prediction. To enhance the image translation ability, we take advantage of a Fully Convolutional Residual Network (FCRN) and combine it with a generative adversarial network, which has shown remarkable achievements in image-to-image tasks. We also present a new loss function including the scale-invariant (SI) error and the structural similarity (SSIM) loss function to improve our model and to output a high-quality depth map. Experiments show that the DepthGAN performs better in monocular depth prediction than the current best method on the NYU Depth v2 dataset.