| KEYWORD |
Area Ingegneria
Convolutional Neural Network cores towards Deep Space
Parole chiave AEROSPACE, ARTIFICIAL NEURAL NETWORKS, CIRCUITS, EMBEDDED SYSTEMS
Riferimenti LUCA STERPONE
Gruppi di ricerca GR-05 - ELECTRONIC CAD & RELIABILITY GROUP - CAD
Tipo tesi RICERCA SPERIMENTALE
Descrizione Nowadays, the widespread adoption of Artificial Intelligence (AI) approaches in embedded systems paven the way of a new generation of computing cores able to outperform standard Huffman based computational approach. The goal of the present thesis is to develop a Neural Network circuit node able to implement Convolutional computation in low power and low area consumption and adopted in Deep Space missions. These missions are characterized by a long duration that is affecting the performances of a device (timing degradation) and consequently the computational capabilities. The thesis is done in collaboration with MicroChip and the European Space Agency.
Scadenza validita proposta 01/04/2021
PROPONI LA TUA CANDIDATURA