Postdoctoral Fellowship

Postdoctoral Fellowship in Deep Learning for Embedded.

Postdoctoral Fellowship in Deep Learning for Embedded and Multimedia IntelligenceQualification

Ph. D. in Computer Science, Electrical Engineering, or related disciplines with a strong background in Artificial Intelligence, Deep Learning, or Computer Vision.

The candidate should have at least two Q1 journal publications (Scopus or Web of Science indexed) and proven expertise in using deep learning frameworks (TensorFlowPyTorch), imagevideo analytics, or real-time edge systems.

The successful candidate will work on AI-driven multimedia research, focusing on video analytics, remote health monitoring, environmental surveillance, and embedded AI.

The position provides an opportunity to collaborate with University of Bourgogne (France), Griffith University (Australia) and SONY India as part of international research initiatives on biomedical signal analysis, computer vision for sustainability, and intelligent monitoring systems.

A key research direction will involve exploring and adapting LeJepa, the new framework proposed by Yann LeCun, for representation and predictive learning in multimodal applications.

Computer Vision and Video Analytics

Remote Health Monitoring using Deep Learning (rPPG, physiological computing)

Deep Learning for Environmental and Wildlife Surveillance

Edge AI for Embedded Vision Systems

AI for Agriculture and Sustainable Development

Representation and Predictive Learning using LeJepa Framework

Responsibilities

Conduct high-quality research aligned with ongoing projects in health monitoring, agriculture, and environmental intelligence.

Publish results in reputed international journals and conferences.

Collaborate with international partners from University of Bourgogne and Griffith University.

Mentor undergraduate and graduate students and support laboratory activities.

Contribute to grant proposal writing and assist in developing funded research initiatives.

Investigate and implement LeJepa-inspired methods for multimodal representation learning.