Reliable forensic evidence based on research, data and good practice.

We currently provide forensic services in facial comparison and case review. Image analysis and enhancement services coming soon.

Casework

  • Facial comparison

    Our innovative approach to facial comparison combines rigorous manual morphological analysis with state of the art open-source facial recognition algorithms, all delivered by experienced forensic experts. We test every aspect of our facial comparison methodology to ensure we provide you with robust, reliable and accurate results.

  • Image analysis

    Coming soon.

  • Image enhancement

    Coming soon

  • Case review

    We provide expert advice on forensic image analysis reports and facial identification evidence, giving you a balanced and impartial review of the evidence. We can also support cold case review relating to images and video.

We are currently working towards accreditation to ISO 17025:2017 and the Forensic Science Regulator’s Codes of Practice. Whilst we are not yet accredited we mitigate the risks of non-compliance with robust method validation, training and competency frameworks, participation in proficiency testing and following documented procedures. Additional information on these mitigations and our accreditation roadmap can be made available on request.

Cases

Here are some of our recent cases that are in the public domain.

  • It all begins with an idea. Maybe you want to launch a business. Maybe you want to turn a hobby into something more.

  • In this case a low resolution version of a video from a political event was used in social media posts to (mis)identify the Trump assassination plot suspect. Journalists were able to source the original 4k resolution video which we could use in a forensic facial examination. The results was then verified with a facial recognition algorithm result. Using the 4k resolution video revealed a high number of dissimilarities in facial feature detail.

    An important example of always trying to obtain the original video/image at its highest quality when carrying forensic facial comparison. Opinions based on compressed and downscaled imagery can lead to inconclusive, and at worst, erroneous conclusions.

    Link to article here.

  • This case was a request for forensic facial comparison to assist in the fact-checking of a social media post on X spreading potential disinformation. Link to the full article here.

    O    nce the images were analysed and compared side-by-side the dissimilarities started to add up, in this case the outer helix of the left ear, inner corner of the eyes, morphology of the mouth and dissimilarities in facial creases (including forehead creases and a crease below the mouth).

    As well as carrying out a robust manual examination, these images were also examined using a facial recognition algorithm for verification. Interestingly, even though the algorithm score was indicative of the images showing different people the score was higher than expected, given the dissimilarities observed in the manual examination. We suspect this was due to the substantial perspective distortion in the first image affecting the geometry of the face (likely caused by the distance of the person from the camera), highlighting the need for trained and competent practitioners to work with and understand automated facial recognition technology, as well as the challenges in reconciling differences in output between manual and automated processes.

Get in touch.