The US National Institute of Standards and Technology (NIST) has published new guidelines it claims will help organizations optimize their efforts to detect face morphing software.
Face morphing is a type of deepfake technology that enables threat actors to blend the photos of two people into a single image. In doing so, it simplifies identity fraud by tricking face recognition systems into erroneously identifying an image as belonging to both original individuals.
In this way, individual A can assume the identity of individual B and vice versa, NIST said.
The new report, Face Analysis Technology Evaluation (FATE) MORPH 4B: Considerations for Implementing Morph Detection in Operations(NISTIR 8584), offers an introduction to the topic and key detection methods.
It focuses mainly on the pros and cons of various investigatory techniques, and ways to prevent morphs from entering operational systems in locations such as passport application offices and border crossings.
Read more on deepfakes:Deepfake AI Trading Scams Target Global Investors
Specifically, it differentiates between two detection scenarios.
In the first, dubbed “single-image morph attack detection,” examiners only have the synthesized photo. In the second, “differential morph attack detection,” they have that photo alongside a second, genuine image.
Pros and Cons
Single-image detection can detect morphs up to 100% of the time at a false detection rate of 1%, but only if the tool has been trained on examples from the software that generated the morph, NIST said. If not, accuracy can fall to under 40%.
Differential detectors are more consistent, with accuracy ranging from 72% to 90% for morphs created via open-source and proprietary software, but they require an additional genuine photo for comparison.
The good news is that detection tools have improved dramatically in recent years, according to report author, Mei Ngan.
“What we’re trying to do is guide operational staff in determining whether there is a need for investigation and what steps that might take,”she added.
“It’s important to know that morphing attacks are happening, and there are ways to mitigate them. The most effective way is to not allow users the opportunity to submit a manipulated photo for an ID credential in the first place.”
As far back as 2022, Europol warned that face morphing could be used in document fraud, such as applying for a passport or passing identity checks.