|
|
|
|
Safety_Critical_Machine_Learning |
|
UNO™ is a safety camera with safe privacy. The camera works
through the background subtraction mechanism. The background is learned at the
time of installation. All people who enter the background are seen as shadows,
and the machine learning model trains a series of positions that represent
imminent danger, a fall, or another accident. The camera does not transmit
images to the outside but only has a Boolean electrical interface that signals
a dangerous situation. In fact, the camera was designed to comply with all
regulations on people's privacy.
The UNO™ camera has been designed using the ZISC® (Zero
Instruction Set Computer®) chip with 78 neurons by Silicon Recognition®. The
chip that today has 1000/5000 neurons is called Neuromem® and is marketed by
General Vision®.
UNO™ has been realized in 2002 as part of a research project
funded by a large company in the security sector. We have now improved the
design by adding network management features and have applied for a patent for
the new version.
|
|
The camera was designed on a
MUREN® (MUltimedia Recognition ENgine) board from Silicon Recognition with
two ZISC78 chips and a plug-in CMOS camera. The design required reconfiguring
the FPGA with new VHDL code. |
|
|
|
General Synaptics Aerospace_and_Defence_Machine_Learning_Company VAT NUMBER:_IT02670700992 REA NUMBER: GE-503104 Email:_luca.marchese@synaptics.org |
|
|
|
Copyright© 2026 General Synaptics |
