Human-centric Artificial Intelligence for Internet of Things: Application to Search and Rescue Operations, 

L’Oréal-UNESCO “For Women in Science” Young Talents 2023, (Budget: 10k Eur, March 2024-February 2025)

Motivation and main objective

The current research on Search and Rescue (SAR) emphasizes the need for integrating diverse technologies to meet various challenges effectively. This proposal introduces a novel three-pronged approach: 1) A human-centric AIoT system that includes wearable sensors, drone modules with acoustic and visual capabilities, multispectral drones, and a ground monitoring center to enhance SAR operations across different terrains and conditions. 2) Real-world experiments with volunteers to gather multimodal data in SAR scenarios, which will be annotated and sampled using human-in-the-loop methods. 3) The development of advanced machine learning and deep learning techniques, focusing on mitigating biases in gender, race, and age in SAR image detection, while ensuring data usage maintains a balance between discrimination mitigation and reliability. This project aims to integrate standardized equipment to generate high-quality multimodal datasets and develop a robust AI framework within a 12-month period, prioritizing responsible and trustworthy AI practices.