|
|
||||
|
|
|
|||
|
|
||||
|
SHAllow Looks One Map (SHALOM™) is a method of REAL
TIME OBJECTS DETECTION that uses SHALLOW NEURAL CLASSIFIERS and CELLULAR AUTOMATA
in order to identify objects in the frames of a video looking only one time
at any single frame. SHALOM™ is inspired by YOLO algorithm but it is based on Shallow Neural
Networks and Cellular Automata. SHALOM™ has been designed to speed up the
identification of specific objects and determine their exact position.
SHALOM™ works with high speed of execution both in the "features
extraction" phase that uses a single image scan without ROS (Region Of
Scanning) and ROI (Region Of Interest), and in the pattern recognition phase
with Shallow Neural Networks on SIMD processors or Neuromem®.
Mythos™ technology enables SHALOM™ to run on Von
Neumann processors like BAE SYSTEMS RAD750™. The Cellular Automaton manages
the behaviour of the fixed grid. |
|
|
|
LUCA MARCHESE Aerospace_&_Defence_Machine_Learning_Company VAT:_IT0267070992 Email:_luca.marchese@synaptics.org |
|
Contacts_and_Social_Media
|
|
Copyright_©_2025_LUCA_MARCHESE |
