EmoP3D: A brain like pyramidal deep neural network for emotion recognition

Incidence of parametrized methods for spectral analysis of Ship Motion

To protect the marine environment and increase the safety of ships and navigation, monitoring the state of sea is fundamental. In current study an accurate analysis of the spectral moments of the heave and pitch motions time series of the S175 containership was carried out, based on the JONSWAP wave spectrum. Two spectrum estimation methods, namely the Periodogram and Thomson ones, have been applied and compared. The performances of both methods are discussed and evaluated carefully.

Breast Cancer Imaging at Microwave Frequencies via Neural Networks

Several biomedical imaging modalities are available in the scientific literature for breast cancer diagnosis, each with its own advantages and limitations. In this framework, microwave tomography has gained a certain interest in the last years due to its capability to provide quantitative maps of tissues electric properties in a safe, non-invasive fashion. Furthermore, this imaging modality proved to be cheaper than traditional medical exams. By virtue of the great advantages that artificial ...

A Novel Authentication System Based on Hand Gesture Ultrasound Signatures

In the framework of security, biometrics proved to be the new frontier for human identification/authentication. It is more and more easy to find system exploiting fingerprint, voice or face recognition in application ranging from forensic to smartphone unlocking. In this paper the possibility of considering hand gesture for authentication purposes is explored. Hand gesture Doppler signatures are extracted by means of an ultrasound prototype and a deep learning based detector is considered. Th...