Brilliant work!Could be totally wrong but gist I get for their Oculi tech from the patent that @Diogenese also found is driven around the pixel identification and changes and they speak of the vision component primarily that needs to dovetail into an AI processing aspect maybe?
Driven around threshold values and trigger signals...presumably on pixel changes?
So, question is for me is, who's AI is doing the backend processing?
Appears they need an AI processor / IP maybe to be in that SPU.
The only reference to AI per se was a comment in the patent and quite generic. Nothing on neural or neuromorphic, SNN, DNN, CNN etc.
Only reference I can find so far to neuromorphic is their site but patent and TM applications etc so far don't show any of their own neuromorphic work so logic indicates sourcing elsewhere....whether us or not is to be formally confirmed, I guess.
![]()
Our Technology-V1 | Oculi
www.oculi.ai
OCULI SPU™: Integrated Neuromorphic Sensing & Processing
![]()
Oculi makes the OCULI SPU™, Sensing and Processing Unit. This is the only Software-Defined Vision Sensor™ and it can deliver real-time vision intelligence (VI) by combining sensing + pre-processing in the pixel.
SENSING AND PROCESSING UNIT GENERATING A TRIGGER SIGNAL UPON OCCURRENCE OF SPECIFIED CONDITIONS - Fris, Inc.
<div p-id="p-0001">A sensing and processing electronic circuit includes a sensing array having a plurality of sensing elements each having a sensor configured to produce an output signal in response twww.freepatentsonline.com
A sensing and processing electronic circuit includes a sensing array having a plurality of sensing elements each having a sensor configured to produce an output signal in response to detection of an environmental parameter. The circuit includes a triggering unit configured to generate a trigger signal that causes an action to occur. The trigger signal is generated based on selected activated sensing elements satisfying one or more specifiable conditions. A sensing element is activated when a characteristic value based on the output value of the sensing element crosses one or more thresholds so that the characteristic value is above, below, within or outside of a specified range of values or when a difference value between the characteristic value and a previous or template value of the sensing element crosses one or more thresholds so that the difference value is above, below, within or outside of a specified range of values.
or it may be more convenient to construct a specialized apparatus or system to perform the required methods and techniques. In some cases the environments in which various embodiments described herein are implemented may employ machine-learning and/or artificial intelligence techniques to perform the required methods and techniques.
They also applied for a more recent TM for a Software-Defined Vision Sensor with a dummy board by the looks so unfortunately no Akida identifier **sad face** haha
Was interesting to see the first use anywhere & first use commerce dates in the application.
Application for Software-defined Vision Sensor
Trademark registration supporting document for Application regarding SOFTWARE-DEFINED VISION SENSOR registered by Fris Incuspto.report
Trademark/Service Mark Application, Principal Register
Serial Number: 97415842 Filing Date: 05/18/2022
To the Commissioner for Trademarks:
MARK: Software-Defined Vision Sensor (Standard Characters, see mark)
The literal element of the mark consists of Software-Defined Vision Sensor. The mark consists of standard characters, without claim to any particular font style, size, or color.
The applicant, Fris Inc, DBA Oculi, a corporation of Delaware, having an address of
5520 Research Park Drive, Suite 100
5520 Research Park Drive, Suite 100
Baltimore, Maryland 21228
United States
410-205-5513(phone)
XXXX
requests registration of the trademark/service mark identified above in the United States Patent and Trademark Office on the Principal Register established by the Act of July 5, 1946 (15 U.S.C. Section 1051 et seq.), as amended, for the following:
For specific filing basis information for each item, you must view the display within the Input Table.
International Class 009: Semi-conductors; Semiconductor chip sets; Semiconductor chips; Semiconductor devices; Semiconductor wafers; Semiconductors; Electronic semi-conductors
In International Class 009, the mark was first used by the applicant or the applicant's related company or licensee or predecessor in interest at least as early as 12/08/2021, and first used in commerce at least as early as 12/10/2021, and is now in use in such commerce. The applicant is submitting one(or more) specimen(s) showing the mark as used in commerce on or in connection with any item in the class of listed goods/services, consisting of aBoard with statement on it: Software-Defined Vision Sensor.
Specimen File1
Specimen File2
Webpage URL: www.oculi.ai
Webpage Date of Access: 12/10/2021
SIGNATURE INFORMATION FEE INFORMATION CORRESPONDENCE INFORMATION GOODS AND/OR SERVICES AND BASIS INFORMATION LEGAL ENTITY INFORMATION APPLICANT INFORMATION MARK INFORMATIONInput Field Entered SERIAL NUMBER97415842 *MARKSoftware-Defined Vision Sensor STANDARD CHARACTERSYES USPTO-GENERATED IMAGEYES LITERAL ELEMENTSoftware-Defined Vision Sensor MARK STATEMENTThe mark consists of standard characters, without claim to any particular font style, size, or color. REGISTERPrincipal *OWNER OF MARKFris Inc DBA/AKA/TA/FormerlyDBA Oculi INTERNAL ADDRESS5520 Research Park Drive, Suite 100 *MAILING ADDRESS5520 Research Park Drive, Suite 100 *CITYBaltimore *STATE
(Required for U.S. applicants)Maryland *COUNTRY/REGION/JURISDICTION/U.S. TERRITORYUnited States *ZIP/POSTAL CODE
(Required for U.S. and certain international addresses)21228 PHONE410-205-5513 *EMAIL ADDRESSXXXX WEBSITE ADDRESSwww.oculi.ai TYPEcorporation STATE/COUNTRY/REGION/JURISDICTION/U.S. TERRITORY OF INCORPORATIONDelaware INTERNATIONAL CLASS009 *IDENTIFICATIONSemi-conductors; Semiconductor chip sets; Semiconductor chips; Semiconductor devices; Semiconductor wafers; Semiconductors; Electronic semi-conductors FILING BASISSECTION 1(a) FIRST USE ANYWHERE DATEAt least as early as 12/08/2021 FIRST USE IN COMMERCE DATEAt least as early as 12/10/2021 SPECIMEN FILE NAME(S)\\TICRS\EXPORT18\IMAGEOUT 18\974\158\97415842\xml1 \ APP0003.JPG \\TICRS\EXPORT18\IMAGEOUT 18\974\158\97415842\xml1 \ APP0004.JPG SPECIMEN DESCRIPTIONBoard with statement on it: Software-Defined Vision Sensor WEBPAGE URLwww.oculi.ai WEBPAGE DATE OF ACCESS12/10/2021 NAMEFris Inc PRIMARY EMAIL ADDRESS FOR CORRESPONDENCEinfo@oculi.ai SECONDARY EMAIL ADDRESS(ES) (COURTESY COPIES)sana.kouatly@oculi.ai; charbel.rizk@oculi.ai APPLICATION FILING OPTIONTEAS Standard NUMBER OF CLASSES1 APPLICATION FOR REGISTRATION PER CLASS350 *TOTAL FEES DUE350 *TOTAL FEES PAID350 SIGNATURENOT PROVIDED SIGNATORY'S NAMENOT PROVIDED SIGNATORY'S POSITIONNOT PROVIDED DATE SIGNEDNOT PROVIDED
View attachment 18496
View attachment 18497
Mar 22 TinyML presso slide.
View attachment 18504
So we need to be careful that we are looking for neuromorphic processing of the signal from the sensor, and not just a bionic event vision sensor.
ie we need just what you have underlined.
But I think that is what they are trying to say with the term "Software-Defined Vision Sensor". ie you change what the vision sensor is looking for by applying SNN processing to the signal. Running that SNN will be in Akida hardware, but changing the parameters, how many layers etc are configurable in software.
Jeeze, one day one of these startups might just tell us straight up: "Akida on-chip" to their specs. If only.