Fine-tuning mug shot match game

MICHIGAN STATE (US) — Newly-developed software is able to automatically match hand-drawn facial sketches to mug shots stored in law enforcement databases.

Typically, artists’ sketches are drawn by artists from information obtained from a witness, but the facial sketch is rarely an accurate depiction of what the suspect looks like.

The few commercial software programs currently available that produce sketches based on a witness’ description tend to be less accurate than sketches drawn by a trained forensic artist.

Details of the technology are reported in the journal IEEE Transactions on Pattern Analysis and Machine Intelligence.

“We’re dealing with the worst of the worst here,” says Brendan Klare, a doctoral student in computer science and engineering at Michigan State University.

“Police sketch artists aren’t called in because someone stole a pack of gum. A lot of time is spent generating these facial sketches so it only makes sense that they are matched with the available technology to catch these criminals.”

“We improved significantly on one of the top commercial face-recognition systems,” Klare says. “Using a database of more than 10,000 mug shot photos, 45 percent of the time we had the correct person.”

All of the sketches used were from real crimes where the criminal was later identified.

“We don’t match them pixel by pixel,” says Anil Jain, professor of computer science and engineering.

“We match them up by finding high-level features from both the sketch and the photo; features such as the structural distribution and the shape of the eyes, nose and chin.”

More news from Michigan State University: