Akademik Veri Yönetim Sistemi
Grad Colloquium'24 Artificial Intelligence, İstanbul, Türkiye, 04 Haziran 2024, ss.20
Finding people in collapsed buildings after disasters is crucial for rescue efforts. Despite the
use of various methods, there are still challenges to ensure accurate and quick rescues. The rapid
detection of living people under debris is crucial. This study aims to introduce a new method:
using a type of artificial intelligence called Temporal Neural Networks (TCN) to find human
respiration under building debris by generating new computer-simulated electromagnetic data
as shown in Figure 1. To achieve this, an electromagnetic simulation program is used to create
a realistic model of building debris and a human. The data is then used for simulating body
movements which are indicating respiration. Measurements are taken between 150 MHz and
650 MHz in different scenarios. The measurements are enriched with different levels of noise,
and different respiration patterns are created. The result is 15 different subsets with a data size
of 2x216.000. The respiration of humans trapped under debris is detected with a success rate of
99.97%. This demonstrates the effectiveness of this method andindicates its potential for further
testing.