Project Details
Description
The Judicial Investigation Agency aims to create a unit within the Police Information Platform office specialized in innovation, research, and development related to data science, artificial intelligence, and machine learning. The purpose is to develop and implement projects that contribute to the fight against organized crime. Below are three priority work scenarios in which they wish to develop AI-based software systems using machine learning and deep learning to combat organized crime from the Judicial Investigation Agency of Costa Rica.
4. Facial recognition in criminal databases - The OIJ (Judicial Investigation Agency) maintains a large collection of images of individuals who have committed crimes and have been part of the Criminal Archives. When a criminal event occurs, the victim must undergo a tedious process of identifying the perpetrator from this enormous collection of photographs. To streamline this process and avoid re-victimization, the aim is to implement automated image recognition, where the program searches for the most similar images upon selection, not only limited to facial recognition but also extending to tattoos, objects, and even videos
5. Automatic categorization of crime reports - Natural language interpretation: Once a crime report is filed, it needs to be assigned a specific category of crime to be directed to a specialized section. Currently, this process is manual, but the intention is to automate it through AI.
6. Crime prediction by location. There is a need to predict potential crime occurrences in different locations across the country. The goal is to identify areas where criminal events are more likely to happen.
This proposal suggests developing three software systems based on Artificial Intelligence, using Machine Learning and Deep Learning, to aid in the fight against organized crime within the Judicial Investigation Agency of Costa Rica. The development of these three software systems will be a collaborative effort between the Technological Institute and the Judicial Investigation Agency of Costa Rica
4. Facial recognition in criminal databases - The OIJ (Judicial Investigation Agency) maintains a large collection of images of individuals who have committed crimes and have been part of the Criminal Archives. When a criminal event occurs, the victim must undergo a tedious process of identifying the perpetrator from this enormous collection of photographs. To streamline this process and avoid re-victimization, the aim is to implement automated image recognition, where the program searches for the most similar images upon selection, not only limited to facial recognition but also extending to tattoos, objects, and even videos
5. Automatic categorization of crime reports - Natural language interpretation: Once a crime report is filed, it needs to be assigned a specific category of crime to be directed to a specialized section. Currently, this process is manual, but the intention is to automate it through AI.
6. Crime prediction by location. There is a need to predict potential crime occurrences in different locations across the country. The goal is to identify areas where criminal events are more likely to happen.
This proposal suggests developing three software systems based on Artificial Intelligence, using Machine Learning and Deep Learning, to aid in the fight against organized crime within the Judicial Investigation Agency of Costa Rica. The development of these three software systems will be a collaborative effort between the Technological Institute and the Judicial Investigation Agency of Costa Rica
General Objective
Contribuir en la lucha contra el crimen organizado mediante sistemas de IA que mejoren la eficiencia y precisión en procesos judiciales.
Research Lines
Inteligencia Artificial
Procesamiento de Lenguaje Natural
Análisis Predictivo
Procesamiento de Lenguaje Natural
Análisis Predictivo
| Status | Active |
|---|---|
| Effective start/end date | 1/01/24 → 31/12/26 |
Keywords
- Deep Learning
- Machine Learning
- IA
- CNN
- GPT
- Time Series
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.