Many people may be familiar with the movie Minority Report. The film tells the story of a technology that predicts the future to prevent crimes before they happen. In the real world, similar technology has already emerged in South Korea through an AI system called Dejaview. This system was developed to predict and prevent crimes by analyzing CCTV footage.
Dejaview: A Technology Combining AI and Machine Learning
Dejaview is the result of research and development by researchers from the Electronics and Telecommunications Research Institute (ETRI) of South Korea. This system uses machine learning technology to analyze CCTV footage. The analysis is based on various data, including the date, time, location, past events, and other key variables, to assess the likelihood of crimes occurring.

How Dejaview works
1. Predicting the behavior of high-risk individuals
The system analyzes the behavior of individualswho are likely to reoffendbased on their pastoffenses. AI surveils the movements and behavior of these individuals in public places, such as train stations or crowded areas, to reduce potential risks.
2. Time and location-based prediction
Dejaview uses historical data to predict high-risk areas and time periods, such as desolate areas at night. The system enables officers to monitor high-risk areas more effectively. According to testing in Seocho, a rural area in South Korea, the system achieved a prediction accuracy of 82.8%.
Dejaview’s AI system was trained on a large dataset collected from over 32,000 surveillance camera video clips, recording various events from the past three years. The system uses this data to analyze and process it, creating a model that can effectively predict crimes in advance.

Debates about AI in crime prevention
Although AI-based crime prevention technology has gained attention and development in many countries, there are still debates regarding data privacy. Many parties are concerned about personal rights violations from CCTV and AI surveillance. However, developers insist that this technology will only be used in necessary areas, such as airports, energy plants, and public safety infrastructure.
Currently, Dejaview is limited to security-critical areas, such as airports, power plants, and national control centers. However, there are plans to expand its commercial use by the end of 2025. This could make Dejaview a key tool for securing public spaces.
In addition to South Korea, Argentina is another example of a country using AI systems for crime prevention. The AI system in Argentina not only analyzes CCTV footage but also uses data from social media, websites, and even the dark web. This helps predict and prevent crimes more comprehensively.
Conclusion
Dejaview exemplifies the use of AI and machine learning technology in crime prevention, enhancing the effectiveness of public security. However, this technology still faces ethical challenges and concerns about data privacy. In the future, if used appropriately, Dejaview could become the new standard for security and stability in societies around the world.
Source: