Back to results
US 12,387,532 B1Granted
Methods and systems for onsite signature verification
AssigneeParascript LLC
InventorsIvan Bulatov, Tatiana Vazioulina
Life of this patent
Estimating the term…
Abstract
A method and system for writable document forgery detection that includes encrypting a reference writing sample, compressing the reference writing sample, embedding the reference writing sample onto the writable document, receiving the writable document, the writable document including a handwritten segment, encrypting the handwritten segment of the received writable document, comparing the encrypted reference writing sample to the encrypted handwritten segment, and assessing a likelihood of the encrypted reference writing sample and the encrypted handwritten segment being written by a same person based on the comparison.
Claims26
A method for writable document forgery detection, the method comprising:
- extracting a reference writing sample from a media via a neural network by generating one or more first embedding vectors, and encoding the one or more first embedding vectors, wherein the one or more first embedding vectors include a biometric characteristic of the reference writing sample generated by the neural network, the biometric characteristic comprising a velocity of the reference writing sample;
- encrypting the one or more first embedding vectors;
- compressing the one or more first embedding vectors;
- printing the one or more first embedding vectors onto the writable document;
- receiving the writable document, the writable document including a handwritten segment;
- extracting one or more second embedding vectors from the handwritten segment of the received writable document;
- comparing the one or more first embedding vectors to the one or more second embedding vectors;
- and
- assessing a likelihood of the reference writing sample and the handwritten segment being written by a same person based on the comparison.
6 more elements
A writable document forgery detection system comprising:
- an updatable data repository;
- a printing device;
- and
- a computing device operatively coupled to the updatable data repository, to the neural network and to the printing device, the computing device comprising a processor including a neural network and a memory, the memory storing instructions that, when executed by the processor, perform a set of operations comprising:
- extracting, via the neural network, a reference writing sample from a media by generating one or more first embedding vectors, and encoding the one or more first embedding vectors, wherein the one or more first embedding vectors include a biometric characteristic of the reference writing sample generated by the neural network, the biometric characteristic comprising a velocity of the reference writing sample;
- encrypting, via the processor, the one or more first embedding vectors;
- compressing, via the processor, the one or more first embedding vectors;
- and
- printing, via the printing device, the one or more first embedding vectors onto the writable document.
6 more elements
A writable document forgery detection system comprising:
- a receiving device;
- an updatable data repository functionally coupled to the receiving device;
- and
- a computing device operatively coupled to the receiving device and to the updatable data repository, the computing device comprising a processor including a neural network and a memory, the memory storing instructions that, when executed by the processor, perform a set of operations comprising:
- receiving the writable document at the receiving device, the writable document including a handwritten segment and one or more first embedding vectors of a reference writing sample printed thereon, wherein the one or more first embedding vectors include a biometric characteristic of the reference writing sample generated by the neural network, the biometric characteristic comprising a velocity of the reference writing sample;
- extracting, via the processor, one or more second embedding vectors from the received handwritten segment of the writable document;
- comparing, via the processor, the one or more first embedding vectors to the-one or more second embedding vectors;
- and
- assessing, via the processor, a likelihood of the reference writing sample and the handwritten segment being written by a same person based on the comparison.
6 more elements
A writable document forgery detection system comprising:
- a receiving device;
- a printing device;
- at least one updatable data repository functionally coupled to the receiving device and to the embedding device;
- at least one computing device operatively coupled to the receiving device and to the at least one updatable data repository, the at least one computing device comprising at least one processor including a neural network and at least one memory, the memory storing instructions that, when executed by the at least one processor, perform a set of operations comprising:
- extracting, via the neural network, a reference writing sample from a media by generating one or more first embedding vectors, and encoding the one or more first embedding vectors, wherein the one or more first embedding vectors include a biometric characteristic of the reference writing sample generated by the neural network, the biometric characteristic comprising a velocity of the reference writing sample;
- encrypting, via the at least one processor, the one or more first embedding vectors;
- compressing, via the at least one processor, the one or more first embedding vectors;
- printing, via the printing device, the one or more first embedding vectors onto the writable document;
- receiving the writable document at the receiving device, the writable document including a handwritten segment;
- extracting, via the at least one processor, one or more second embedding vectors from the handwritten segment of the writable document;
- comparing, via the at least one processor, the one or more first embedding vectors to the one or more second embedding vectors;
- and
- assessing, via the at least one processor, a likelihood of the reference writing sample and the handwritten segment being written by a same person based on the comparison.
10 more elements
Dependencies are inferred from the claim text.