AI crime prevention in urban areas

Issue 1 2022 News & Events

Identifying location-specific attributes is an important aspect of social artificial intelligence. However, models that are frequently trained on subjective perceptions and still images are unreliable in predicting crime. Now researchers from GIST in Korea take things to the next level by training a neural network with a geo-tagged dataset of reported deviant incidents and sequential images of deviant locations to accurately determine unsafe locations by linking the deviant behaviour to the visual features of a city.

Identifying possible hotspots of crime in a city is an important issue for urban safety development and can help the authorities take necessary steps to make the city safer for its residents. The effectiveness of such preventive measures depends on the accuracy of the predictions, which are increasingly being made by artificial intelligence (AI)-based models. Most existing models use subjective perceptions of safe locations, socioeconomic status and still images of crime scenes and only a few violent crimes are categorised as input data. As a result, there is often a discrepancy between their predictions and reality.

In a new study published in AAAI Conference on Artificial Intelligence, researchers from the Gwangju Institute of Science and Technology (GIST) in South Korea proposed a different strategy based on a large-scale dataset and the concept of 'deviance', which included not only violent crimes but also civil complaints regarding behaviours violating social norms, which is also called 'deviant behaviour'.

Accordingly, they developed a convolutional neural network model, aptly called DevianceNet and trained it using a geo-tagged dataset of deviant incident reports with corresponding sequential images of the incident locations acquired using Google street view. “Our work is the first study that investigates the relationship between the physical appearance of a city and deviance with deep learning techniques,” comments associate professor Hae-Gon Jeon, who headed the study.

The researchers collected the images from 10 GPS coordinates within a radius of 50 m from the site of reported incidents and for each GPS location, considered images with 12 directions for a total of 120 images. Using data from five major cities in South Korea and two in the USA, they trained and tested their model with 2250 deviant places and 760 952 images. Such a large dataset enhanced the prediction capabilities of the model to detect possible deviant locations.

“This improved visual perception tasks such as recognition, classification and localisation,” explains Jeon. “The holistic representation of DevianceNet extracted from entire image sequences makes it possible to accurately classify and detect deviant places.

“Since the model can identify deviant behaviour from the visual attributes of the environment, it is not city-specific and can be used to identify potential unsafe locations even when criminal incident data is not available. “This makes it a useful tool in countries that have poor record keeping. The model can also be integrated into navigational services to suggest safer routes,” says Jeon, speaking on the practical implications of the study.

“Additionally, city planners can use the results of the prediction to understand how the city’s layout or design environment can be redesigned to lower instances of deviant behaviour and criminal activity.”

Find out more at, or

Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

Pentagon appointed as Milestone distributor
Elvey Security Technologies News & Events Surveillance
Milestone Systems appointed Pentagon Distribution (an Elvey Group company within the Hudaco Group of Companies) as a distributor. XProtect’s open architecture means no lock-in and the ability to customise the connected video solution that will accomplish the job.

SMART Estate Security returns to KZN
Nemtek Electric Fencing Products Technews Publishing Axis Communications SA OneSpace Editor's Choice News & Events Integrated Solutions IoT & Automation
The second SMART Estate Security Conference of 2024 was held in May in KwaZulu-Natal at the Mount Edgecombe Estate Conference Centre, which is located on the Estate’s pristine golf course.

From the editor's desk: Just gooi a cable
Technews Publishing News & Events
      Welcome to the 2024 edition of the SMART Estate Security Handbook. We focus on a host of topics, and this year’s issue also has a larger-than-normal Product Showcase section. Perhaps the vendors are ...

Secutel wins OSPA Award for Outstanding New Security Product
Secutel Technologies News & Events Access Control & Identity Management
[Sponsored] Secutel Technologies’ NoKey Access Control solution won the Outstanding New Security Product category at the 2024 OSPAs in South Africa. The awards were presented at Securex 2024 where all category finalists were recognised for their contribution to the security industry.

ONVIF launches new working groups for cloud, metadata and audio
News & Events Surveillance
ONVIF, the global standardisation initiative for IP-based physical security products, is announcing the formation of three new working groups to tackle standardisation work in cloud connectivity, audio, and advanced metadata.

Inaugural Gallagher Security Johannesburg networking roadshow
Gallagher News & Events
Held at Johannesburg’s Foghound Coffee Company in Midrand from 11 to 12 June, security industry professionals gather at the inaugural Gallagher Security Johannesburg Networking Roadshow.

Trend Micro launches first security solutions for consumer AI PCs
Information Security News & Events
Trend Micro unveiled its first consumer security solutions tailored to safeguard against emerging threats in the era of AI PCs. Trend will bring these advanced capabilities to consumers in late 2024.

Dallmeier receives ISO 27001 certification
Dallmeier Electronic Southern Africa Surveillance News & Events
Dallmeier has received ISO 27001 certification for its Information Security Management System (ISMS). The international standard for information security management ensures that companies meet the highest standards of data protection and data security.

Local manufacturing is still on the rise
Hissco Editor's Choice News & Events Security Services & Risk Management
HISSCO International, Africa's largest manufacturer of security X-ray products, has recently secured a multi-continental contract to supply over 55 baggage X-ray screening systems in 10 countries.

SAIDSA achieves ISO 9001 certification
SAIDSA(SA Intruder Detection Services Association) Associations News & Events Training & Education
The South African Intruder Detection Services Association (SAIDSA) has announced that it has achieved ISO 9001:2015 certification. This milestone reflects its commitment to quality management and excellence in the security services industry.