Research Identity

I am a Postdoctoral Researcher at Wageningen University & Research (WUR), where my work focuses on advancing Self-Supervised Learning (SSL) and Representation Learning for complex, multimodal, and spatiotemporal data.

My research addresses the challenge of extracting meaningful information from large-scale unlabeled datasets, with a particular emphasis on Remote Sensing and Earth Observation. I develop methods that leverage temporal dependencies and cross-modal relationships to build robust, generalizable foundation models for environmental monitoring, agriculture, and climate science.

Research Interests

• Self-Supervised Learning: Contrastive learning, masked modeling, and generative pre-training for non-textual data.
• Multimodal Learning: Integration of multispectral, temporal, and spatial information in unified representation spaces.
• Spatio-Temporal Modeling: Capturing long-range dependencies in video and environmental time series.
• Remote Sensing: Processing satellite imagery for land-use classification and change detection.
• Representation Learning: Enabling autonomous discovery of structure in diverse scientific datasets.
• Foundation Models: Developing scalable, domain-agnostic models for Earth observation tasks.