Deep Mapping from Thermal-to-Visible for Night-Time Face Recognition
In this project, we investigate a deep coupled learning framework to address the problem of matching non-visible face photos against a gallery of visible faces. The coupled framework contains two sub-networks one dedicated to the visible spectrum and the second sub-network dedicated to the non-visible spectrum, as shown in Figure 1. Each sub-network consists of a generative adversarial network architecture. Inspired by the dense network which is capable of maximizing the information flow among features at different levels, we utilize a densely connected encoder-decoder structure as the generator in each GAN sub-network. The coupled GAN framework will be optimized using multiple loss functions. We propose a coupled deep neural network architecture which forces the features from each sub-network to be as close as possible for the same classes in a common latent subspace, while simultaneously preserving information from the input space. To achieve the realistic photo reconstruction while preserving the discriminative information.