Kerala police may soon use US tool to identify online pedophiles through their knuckles
She said that the system is still in the experimental stage but has shown promising early results.
She said that the system is still in the experimental stage but has shown promising early results.
She said that the system is still in the experimental stage but has shown promising early results.
Kochi: In a bid to strengthen the fight against online child sexual abuse, the Kerala Police are in talks to adopt a groundbreaking forensic tool from the United States that can help identify perpetrators from nothing more than their knuckles and fingernails.
Developed by researchers at Michigan State University (MSU), the ‘Suspect Hand Biometrics Tool’ uses artificial intelligence to compare the unique features of a person’s knuckle creases and fingernail beds captured in Child Sexual Abuse Material (CSAM) with publicly available images of potential suspects. According to Professor Kate Seigfried-Spellar, a forensic analyst at MSU, the approach is based on the finding that these areas are unique to every individual, like fingerprints.
“Offenders in CSAM videos are often careful not to show their faces, but they cannot always hide their hands,” said Seigfried-Spellar at c0c0n 2025, the Kerala Police’s annual cybersecurity conference in Kochi on Saturday.
“Our tool focuses on the top of the hand, particularly the knuckles and fingernail beds. It automatically crops those areas from an image, analyses their features, and provides a likelihood of a match that investigators can then manually verify. Other unique biometrics like vein recognition of the hand, moles, and scars can also be analysed,” she said.
She said that the system is still in the experimental stage but has shown promising early results. “We have started testing the tool on actual CSAM material, and our initial accuracy rate is around 95%. Our tool automates the whole process of knuckle and fingernail verification. It is not meant to replace manual verification but to make the process faster and more reliable,” Seigfried-Spellar said.
She added that the research team collaborates closely with law enforcement agencies, particularly the Internet Crimes Against Children Task Force in the US, and intends to make the tool freely available to police departments worldwide once fully developed. The tool could also help identify whether the same perpetrator appears in multiple CSAM videos, a breakthrough that would allow investigators to link related cases more efficiently.
“We have initiated talks with Kerala Police. They have invited me to present the tool at cOcOn,” she said.
IG P Prakash, who oversees the state’s cybercrime division, confirmed that the state police have expressed interest in being part of the testing phase of the tool. “They have promised to give it to us once it’s ready. It is a very interesting proposal, and I have asked them to involve us when the testing begins,” he said.
He added that such tools are crucial to help investigators identify offenders who deliberately conceal their faces. “We already have systems that focus on victim identification, but we also need advanced tools to identify offenders, especially in online child sexual abuse cases,” Prakash said.
Prakash also said that the police plan to set up a dedicated database, similar to the ones maintained by the CBI and Interpol, to store and analyse data on identified offenders.