This project aims to design and study a social robot specifically for use in the medical field, especially within hospital environments. It will provide companionship, emotional support, and assistance, enriching the experience of both patients and healthcare professionals.

I focused specifically on researching and developing the robot's vision system, with a particular emphasis on soft biometric recognition. My primary responsibilities included conducting a comprehensive literature review, evaluating existing approaches, and innovating deep learning models capable of recognizing gender and estimating the age of individuals wearing medical masks.

Leveraging advanced unsupervised domain adaptation techniques, I developed a PyTorch-based deep learning model that outperformed existing state-of-the-art models by achieving a 5% improvement in gender recognition accuracy on individuals wearing medical masks. Inspired by MobileNet and Adversarial Complementary Learning, this model showcases the potential of innovative solutions in enhancing biometric recognition performance.