CMOS video cameras with high dynamic range (HDR) output are particularly suitable for driving assistance applications, where lighting conditions can strongly vary, going from direct sunlight to dark areas in tunnels. However, common visualization devices can only handle a low dynamic range, and thus a dynamic range reduction is needed. Many algorithms have been proposed in the literature to reduce the dynamic range of still pictures. Anyway, extending the available methods to video is not straightforward, due to the peculiar nature of video data. We propose an algorithm for both reducing the dynamic range of video sequences and enhancing its appearance, thus improving visual quality and reducing temporal artifacts. We also provide an optimized version of our algorithm for a viable hardware implementation on an FPGA. The feasibility of this implementation is demonstrated by means of a case study.

This paper proposes a new neural network structure for image processing whose convolutional layers, instead of using kernels with fixed coefficients, use space-variant coefficients. The adoption of this strategy allows the system to adapt its behavior according to the spatial characteristics of the input data. This type of layers performs, as we demonstrate, a non-linear transfer function. The features generated by these layers, compared to the ones generated by canonical CNN layers, are more complex and more suitable to fit to the local characteristics of the images. Networks composed by these non-linear layers offer performance comparable with or superior to the ones which use canonical Convolutional Networks, using fewer layers and a significantly lower number of features. Several applications of these newly conceived networks to classical image-processing problems are analyzed. In particular, we consider: Single-Image Super-Resolution (SISR), Edge-Preserving Smoothing (EPS), Noise Removal (NR), and JPEG artifacts removal (JAR).

Face recognition functions are today exploited through biometric sensors in many applications, from extended security systems to inclusion devices; deep neural network methods are reaching in this field stunning performances. The main limitation of the deep learning approach is an inconvenient relation between the accuracy of the results and the needed computing power. When a personal device is employed, in particular, many algorithms require a cloud computing approach to achieve the expected performances; other algorithms adopt models that are simple by design. A third viable option consists of model (oracle) distillation. This is the most intriguing among the compression techniques since it permits to devise of the minimal structure that will enforce the same I/O relation as the original model. In this paper, a distillation technique is applied to a complex model, enabling the introduction of fast state-of-the-art recognition capabilities on a low-end hardware face recognition sensor module. Two distilled models are presented in this contribution: the former can be directly used in place of the original oracle, while the latter incarnates better the end-to-end approach, removing the need for a separate alignment procedure. The presented biometric systems are examined on the two problems of face verification and face recognition in an open set by using well-agreed training/testing methodologies and datasets.