Choënne Straatman (26) has already completed a master’s degree in Forensic Science at the University of Amsterdam. She also works as a consultant to the district courts in Maastricht and Roermond. How strong is DNA evidence in a particular case? How do DNA traces compare to the forensic evidence? And what kinds of questions might the judge ask the defence about it?
To increase her knowledge of criminal law, Straatman recently completed a master’s degree in Forensics, Criminology and Law at UM. She zeroed in on facial comparison and facial recognition in her thesis Even lachen naar het vogeltje! (Smile for the camera!).
In the Netherlands, facial comparison analysis is performed by the Netherlands Forensic Institute (NFI) in The Hague. It involves comparing a recent photo of a suspect to an image captured by a security camera, for example. The National Police Corps also compares faces using a system called CATCH. It uses an algorithm that runs police photographs through a database of existing face images. This results in a “catch”, a list of possible suspects.
Facial comparison differs from facial recognition, explains Straatman. Facial recognition is when police officers recognise a suspect because they have seen him before. Judges take this kind of evidence very seriously. In 2018, the Amsterdam Court of Appeal stated that, “[…] Faces are unique in nature and most people are excellent at recognising faces.”
According to Straatman, this is quite reductive – especially in light of the literature on this topic. “It matters a lot whether the officer knows a burglar who was caught on a security camera from a single arrest, or if he sees him around the neighbourhood every week. In the latter case, the officer is usually right. But not if he arrested the suspect once, and certainly not if the suspect was one of thirty different people the officer arrested that day.
Straatman is of the opinion that judges should ask more questions about this type of facial recognition evidence, including police photographs and camera images. Has the officer seen the images several times? Were they displayed daily on a monitor in the station canteen, for example? Was the case discussed among colleagues and, if so, did anyone happen to mention local petty thief Harry?
If so, it’s more likely that the officer was biased. The images and conversations may have influenced his judgment, making him believe he recognised Harry. “Judges don’t always ask questions about this. They don’t want it to look like they’re doubting the officer’s professional competence. But that shouldn’t even be a concern, because confirmation bias – as this is called – is unconscious.”
Straatman would like to start a public debate about this topic. That’s why she is currently converting parts of her thesis into an article she would like to submit to a Dutch legal journal like Nederlands Juristenblad.