Organizers

Prof. Mihai Ivanovici,
Transilvania University of Brasov, Romania

Prof. Corneliu Florea,
National University of Science and Technology POLITEHNCA of Bucharest

Description

Smart agriculture relies on Artificial Intelligence (AI) models and Earth Observation (EO) data. EO data is Big Data, considering only the peta-bytes per day from the Copernicus EO programme of EU. In this context, the AI models should learn from/deal with multimodal data. The development of robust AI models for smart agriculture fundamentally depends on access to large-scale, high-quality EO datasets that capture the spatial, temporal and spectral variability inherent in agricultural systems. Using Deep Learning architectures for crop classification, phenological monitoring and stress detection requires massive volumes of data spanning diverse geographical regions, climatic conditions and growing seasons to ensure model generalizability.

Multimodal data fusion for agriculture represents an emerging interdisciplinary field that combines heterogeneous data sources—including satellite imagery, drone-based sensors, IoT devices, weather data, soil sensors, and genomic information—to create comprehensive analytical frameworks for precision agriculture and sustainable food production. This approach leverages advanced machine learning, computer vision and signal processing techniques to integrate temporal, spatial and spectral data streams, enabling more accurate crop monitoring, yield prediction, disease detection and resource optimization than any single data modality could achieve alone.

The field addresses critical challenges in feeding a growing global population while minimizing environmental impact, drawing on expertise from remote sensing, agricultural science, data science and environmental engineering.

The field addresses critical challenges in feeding a growing global population while minimizing environmental impact, drawing on expertise from remote sensing, agricultural science, data science, and environmental engineering.

This special session welcomes contributions that uses different types of data for agricultural applications and are related to one or more topics of interest:

Fusion Methodologies and Algorithms

1. Deep learning architectures for multimodal integration
2. Spatiotemporal data fusion techniques
3. Uncertainty quantification in fused predictions
4. Transfer learning across different sensors and geographic regions
5. Attention mechanisms for modality weighting
6. Multimodal large models for agriculture

Data Acquisition and Sensing Technologies

1. Hyperspectral and multispectral imaging systems
2. UAV-based agricultural monitoring
3. IoT sensor networks for soil moisture, nutrients, and microclimate
4. Synthetic aperture radar (SAR) for all-weather crop monitoring
5. Phenotyping platforms and high-throughput field sensing

Agricultural Applications

1. Crop yield forecasting and early warning systems
2. Plant disease and pest detection
3. Precision irrigation and fertilization management
4. Soil health assessment and carbon sequestration monitoring
5. Livestock monitoring through integrated sensor systems

Emerging Challenges

1. Edge computing and on-farm processing
2. Data standardization and interoperability across platforms
3. Climate adaptation and resilience modeling

Integration with Decision Support

1. Explainable AI for farmer-facing applications
2. Real-time alert systems and recommendation engines
3. Economic modeling and cost-benefit analysis
4. Policy implications and regulatory frameworks

Technical Program Committee

Prof. Erchan Aptoula, Sabanci University, Turkey
Prof. Jacques Blanc-Talon, DGA/DT/TA/EMOO, France
Prof. Fabio del Frate, Università di Roma Tor Vergata, Italy
Prof. Josiane Mothe, UT2J, Toulouse, France

Key Dates

Paper submission

Details about template and submission in
Author Guidelines

Submit your paper via ConfTool:

Submit Paper