In-car Child Abandonment Detection

This abstract presents a method for child detection in car interiors using a Single-Shot Detector (SSD). The objective is to automatically identify the presence of a child within a car for enhanced safety and welfare. The SSD framework, a state-of-the-art object detection algorithm, is trained on a dataset containing annotated images of car interiors with and without children. The model learns to recognize and localize the specific features associated with a child's presence. Images or video frames of the car interior are processed by the trained SSD model to detect and localize child instances accurately. The system offers real-time detection, accurate localization, and potential integration with existing car monitoring systems or applications. Evaluation involves benchmarking performance on various car interior images and assessing detection accuracy. The proposed system aims to contribute to child safety by preventing incidents associated with leaving children unattended in vehicles. It provides a robust and reliable solution for automatically detecting child presence in cars, mitigating risks like vehicular heatstroke, and promoting child welfare.