Smartwatch App to verify signatures developed by Israeli Researchers

Signature verification is a tricky thing. It's often seen that unless a person is very familiar with a signature, he/she is not able to distinguish between a genuine and a forged one. It needs trained human eyes to distinguish between a genuine and a forged signature. There's an opposite side to signature verification as well. Sometimes when the detection of forged signatures is good, there're always the chances of having a case of false positive. That's the signature is not forged and is
Smartwatch App to verify signatures by using motion sensors in smartwatch developed by Israeli Researchers
rendered by the correct person; but due to some human impulses such as palpitation, perspiration or; due to the uncomfortable writing instruments and surfaces (such as a tablet, smartphone or other devices), the signature strokes change dramatically.

This makes signature verification tricky.

That said, we may be getting into the habit of signing perfectly on digital surfaces, but still the accuracy of strokes will come with time. This makes we susceptible to Financial frauds linked to forged signatures.

Signature verification technologies therefore are extremely important in online finance and e-commerce.

Israeli researchers, from Ben-Gurion University of the Negev (BGU) and Tel Aviv University (TAU), have developed software that uses motion data gathered from the movements of a person’s wrist to identify the writer during the signing process.

The researchers' software works by sensing the motion through the motion sensors in present day smartwatches, activity wrist bands etc. The software does two things : It senses changes in rotational motion and orientation of the wrist while making a signature, using accelerometer and gyroscope sensors present in the smartwatches; then the software verifies the signature from its database having multiple signatures made by the same person.

Sicne the software system trains a machine-learning algorithm to distinguish between genuine and forged signatures, hence over a short period of time time, the system is able to verify handwritten signatures and detect even the most skilled forgeries.

How the new Signature Forgery detection system works

The person has to wear the motion sensing device on the wrist of the hand, he/she signs with.

The software tracks the motion of the wrist while signing. It then compares the multiple signature copies rendered by the said person from its database.

Since the system uses machine learning artificial intelligence, hence it detects the genuine signature within a time window of a few minutes, if not seconds.

The system is novel as it can be used with the signature rendered with a traditional pen, on the most commonly used surface, Paper. The app sees a market for it as at present , 1 in every 6 people on this planet (around 16 percent) own a smartwatch.

The researchers plan to make the system more accurate by incorporating the biological data tracked by the wrist trackers.


No comments