If you have an interest in actual violence, you may be familiar with the frequent occurrence of the wrong individual being identified in a police line-up. A group of psychologists suggests that the reason behind this issue lies in the limitations of traditional police line-up procedures, which often hinder witnesses from accurately selecting suspects. They propose that new AI technology could significantly aid in jogging witnesses’ memories, thereby enhancing the identification process.
Contrary to the portrayal in Hollywood movies, modern police line-ups do not involve victims selecting suspects from a physical lineup. Instead, victims are presented with static images, offering limited insights into potential offenders. To address this limitation, experts from the University of Birmingham have developed an innovative interactive program. This program showcases perpetrators through dynamic 3D models from various angles, including the perspective initially observed by the witness.
The research findings indicate that interactive viewing enhances witness selection accuracy by 42%. Even when suspects are presented in video line-ups with rotational views, the technology still boosts selection accuracy by 20%. Professor Heather Flowe, leading the project for over nine years, emphasizes the importance of improving identification accuracy for both innocent individuals and actual perpetrators while revamping existing line-up procedures.
Flowe criticizes the inadequacy and unreliability of current line-up methods employed by law enforcement in the UK and the USA. She highlights the inconsistency in suspect images, ranging from driver’s license photos to individuals in prison attire, leading to a disjointed line-up process. By incorporating AI technology, Flowe’s team aims to introduce memory-jogging techniques such as active physical movements, varied facial expressions, lighting changes, and accessories like masks. The goal is to enhance witness recall and selection accuracy by providing a more comprehensive array of cues.
The innovative line-up system allows witnesses to interact with suspect images, enabling them to manipulate views in a 3D environment. This interactive approach enables witnesses to examine suspects from multiple perspectives, offering a more holistic visual experience. The interdisciplinary research team, comprising experts in psychology, computational modeling, and artificial intelligence, leveraged memory and facial recognition research along with the UK police force’s extensive video database to develop these interactive tools.
Flowe envisions this as the “ultimate line-up technique,” empowering witnesses to retrace their observations and make more informed identifications. By citing cases like that of Andrew Malkinson, wrongfully convicted due to flawed identification, Flowe underscores the significance of improving line-up practices for fairer justice outcomes. She emphasizes the need to consider witnesses’ perspectives and aims to enhance suspect portrayal by incorporating emotional expressions in line-ups.
As technology advances, Flowe anticipates that AI will enable the creation of more realistic suspect images, potentially revolutionizing traditional line-up procedures. Ongoing research will explore the impact of displaying emotional cues on suspects’ faces during line-ups, with collaboration from international research institutions. Despite challenges in eliciting genuine emotional expressions from suspects, AI-driven solutions offer promising avenues for enhancing the accuracy and fairness of identification processes.
