Traces AI Developing Computer Vision-Based Facial Recognition Alternative

A new Y Combinator startup called Traces AI is working on a surveillance alternative to facial recognition. The computer vision solution tracks people moving through crowds based on a number of other visual factors.

“It’s a combination of different parameters,” explained Traces AI Co-Founder Veronika Yurchuk. “We can use your hair style, whether you have a backpack, your type of shoes and the combination of your clothing.”

“We are trying to propose an alternative that will be very effective but less invasive of privacy,” added Co-Founder Kostya Shysh.

The Traces platform intentionally blurs people’s faces before analyzing the data, partly to ensure privacy and partly to guard against the racial and gender biases that have been observed in some facial recognition solutions.

Of course, the system does have certain limitations. It would have trouble tracking people over the course of several days, or if the subject takes off their jacket and puts on a different hat. However, it is robust enough to locate people in a closed environment with virtually no prior information, such as a lost child wearing a blue t-shirt in a shopping mall, and it could do a better job of following someone wearing a mask to avoid detection.

It’s also far less invasive than facial recognition, to the point that it could potentially get around a facial recognition ban like the one recently implemented in San Francisco. That would allow cities to keep an eye on suspicious individuals while still respecting the privacy of law abiding citizens, which could be appealing as smart city solutions become more popular.

Source: Tech Crunch

(Originally posted on Mobile ID World)