PhD in Generative AI for Mobile and Digital health

Social network you want to login/join with: PhD in Generative AI for Mobile and Digital health, pisa col-narrow-left Client: IIT-CNR Location: Job Category: Other - EU work permit required: Yes col-narrow-right Job Reference: 90414664365637632033710 Job Views: 4 Posted: 26.04.2025 Expiry Date: 10.06.2025 col-wide Job Description: The Ubiquitous Internet Unit of IIT-CNR (Pisa, Italy) is seeking highly motivated university graduates (or that will graduate by next October 2025) with a strong academic background who are interested in pursuing a PhD in Generative AI for Mobile and Digital health . Candidate profile The candidate must have (or be about to obtain) a Master’s degree (or equivalent) in Computer Science, Computer Engineering, Biomedical Engineering, Mathematics, or a closely related field. An excellent academic record is required, with outstanding marks in subjects such as machine learning and artificial intelligence. Technical Skills The following skills are an advantage for the candidate: Experience with ML and deep learning frameworks Proficiency in Python PhD topic The generation of synthetic mobile sensor data is a critical challenge for advancing machine learning applications in health monitoring, human activity recognition, and edge AI systems. Traditional approaches, such as GANs and VAEs, have demonstrated success in generating realistic sensor data but often struggle with temporal consistency, multimodal coherence, and personalization. Recent advancements in Large Language Models (LLMs) offer new opportunities for generative tasks, yet their ability to model structured time-series data remains under-explored. The research challenges addressed by this PhD include (but are not limited to) the investigation of LLMs potentialities in the generation of synthetic multimodal sensor data, exploiting also the possible integration with other generative models (GANs, Diffusion Models, and Transformer-based architectures) to enhance realism, efficiency, and privacy. The study will explore: (1) baseline performances of LLMs on sensor data generation (2) new solutions to improve temporal consistency and multimodal integration (3) optimization of model efficiency for on-device and federated deployment (4) privacy preservation The study plan will include the validation of the synthetic sensor data in real-world machine learning applications in the m-health domain, such as stress detection, physical activity recognition, and digital health analytics. J-18808-Ljbffr

Location: provincia-di-pisa, IT

Posted Date: 4/30/2025