Tongyeong City Prevents 'Extreme Choices' Using AI Deep Learning Technology

CCTV Suicide Prevention System Operated at Chungmu and Tongyeong Bridges

A suicide attempter on Chungmudaegyo Bridge captured by the suicide prevention system CCTV in Tongyeong-si, Gyeongnam.

[Asia Economy Yeongnam Reporting Headquarters Reporter Lee Seryeong] A citizen who attempted to jump off Chungmudaegyo Bridge in Tongyeong City, Gyeongnam, was rescued by police who responded after hearing the CCTV suicide prevention system alarm.

According to Tongyeong City on the 3rd, Mr. A, in his 40s, who wandered around Chungmudaegyo Bridge once around midnight on the 1st and left, returned about 20 minutes later and tried to jump by climbing over the bridge railing.

The police received a suicide risk alert issued by the suicide prevention system and rushed to the scene to save Mr. A.

The suicide prevention system is a city-specialized project promoted when establishing the smart city integrated platform in 2021 and has been operating on Chungmudaegyo Bridge and Tongyeong Bridge since April last year.

Based on artificial intelligence (AI) deep learning technology, the system learns and analyzes the behavior characteristic of jump attemptors, specifically the ‘lean’ behavior of leaning their back against the railing, to detect suicide risk in advance.

Deep learning is known as a representative technology in the field of AI representation learning, a deep machine learning technique that researches methods for machines to extract useful information from collected data based on artificial neural networks.

According to the city, more than 70 risk alerts have sounded so far, and there have been three reports at Chungmudaegyo Bridge alone in the past month.

The city stated that it has been continuously analyzing data on basic human behaviors and risky actors by removing noise such as background shaking and screen distortion from CCTV footage of Chungmudaegyo Bridge and Tongyeong Bridge.

Recently, to increase the compiling speed that converts CCTV video analysis and computer language code, the city expanded the application of the ‘TensorFlow Model’ version in the deep learning technique to shorten object analysis time and improve accuracy.

A city official said, “We are striving to prevent suicides using convergent AI technology to overcome the dishonor of having the highest suicide rate among OECD countries and to prevent the misfortune of families of suicide victims,” adding, “We will analyze and apply various patterns such as the behavioral characteristics of suicide attempters to advance related systems and help prevent extreme choices.”

This content was produced with the assistance of AI translation services.