Dhafer Ferchichi
Research Subjects
Dhafer Ferchichi has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Manuel A. Matias
Instituto de Física Interdisciplinar y Sistemas Complejos (IFISC, CSIC-UIB), Campus UIB, 07122 Palma de Mallorca, Spain
2 shared publications - Pablo Moreno-Spiegelberg
IFISC (CSIC-UIB), Instituto de Física Interdisciplinar y Sistemas Complejos, E-07122 Palma de Mallorca, Spain
2 shared publications - Tomàs Sintes
Instituto de Física Interdisciplinar y Sistemas Complejos, IFISC (CSIC-UIB), Universitat de les Illes Balears, E-07122 Palma de Mallorca, Spain
2 shared publications - Àlex Giménez-Romero
Department of Ecology and Complexity, Centre d’Estudis Avançats de Blanes (CEAB-CSIC), Blanes, Girona, Spain
2 shared publications
Latest Publications
- Open Access
- 1 November 2025
- Published by Elsevier in Ecological Indicators
- Vol. 180
A generalizable deep learning framework for large-scale mapping of seagrass habitats
- 1 November 2025
- Published by Elsevier in Ecological Indicators
- Vol. 180
Open AccessA generalizable deep learning framework for large-scale mapping of seagrass habitats
- Open Access
Mapping the distribution of seagrass meadows from space with deep convolutional neural networks
Seagrass meadows play a vital role in supporting coastal communities by promoting biodiversity, mitigating coastal erosion and contributing to local economies. These ecosystems face significant threats, including habitat loss and degradation or climate change. United Nations has recognized the urgency of conserving marine ecosystems, highlighting the need for evidence-based conservation strategies and high-quality monitoring. However, traditional monitoring approaches are often time-consuming, labor-intensive, and costly, limiting their scalability and effectiveness. The growing availability of remote sensing data coupled to the rise of machine learning technologies offer an unprecedented opportunity to develop autonomous, efficient and scalable monitoring systems. Despite many efforts, the development of such systems for seagrass meadows remains a challenge, with recent attempts presenting several limitations such as limited satellite imagery, inadequate metrics for evaluating model performance or insufficient ground truth data, leading to simple proof of concepts rather than useful solutions. Here, we overcome these limitations by developing a comprehensive framework to map Posidonia oceanica meadows in the Mediterranean Sea using an extensive georeferenced habitat dataset and diverse satellite imagery for model training. We successfully evaluate the model generalization capability across different regions and provide the trained model for broader application in biodiversity monitoring and management.
Open AccessMapping the distribution of seagrass meadows from space with deep convolutional neural networks
Seagrass meadows play a vital role in supporting coastal communities by promoting biodiversity, mitigating coastal erosion and contributing to local economies. These ecosystems face significant threats, including habitat loss and degradation or climate change. United Nations has recognized the urgency of conserving marine ecosystems, highlighting the need for evidence-based conservation strategies and high-quality monitoring. However, traditional monitoring approaches are often time-consuming, labor-intensive, and costly, limiting their scalability and effectiveness. The growing availability of remote sensing data coupled to the rise of machine learning technologies offer an unprecedented opportunity to develop autonomous, efficient and scalable monitoring systems. Despite many efforts, the development of such systems for seagrass meadows remains a challenge, with recent attempts presenting several limitations such as limited satellite imagery, inadequate metrics for evaluating model performance or insufficient ground truth data, leading to simple proof of concepts rather than useful solutions. Here, we overcome these limitations by developing a comprehensive framework to map Posidonia oceanica meadows in the Mediterranean Sea using an extensive georeferenced habitat dataset and diverse satellite imagery for model training. We successfully evaluate the model generalization capability across different regions and provide the trained model for broader application in biodiversity monitoring and management.