Adel Hafiane

Artificial intelligence in agricultural remote sensing requires a shift from purely empirical deep learning toward mechanistically grounded, physics-informed models. Although empirical deep learning achieves high accuracy on local datasets, its inability to incorporate explicit biophysical constraints limits its generalization across diverse agricultural environments. This paper reviews the application of Physics-Informed Machine Learning, specifically Physics-Informed Neural Networks in crop monitoring. Furthermore, this study highlights how embedding fundamental biophysical laws, such as Radiative Transfer Models and Surface Energy Balance equations, directly into the neural network optimization process regularizes the learning space. This hybrid approach restricts the solution space of satellite data inversion to physically permissible states, reduces the reliance on large labeled field datasets and ensures physically consistent predictions of agronomic variables, such as soil moisture and crop water stress. Finally, current limitations are outlined, presenting an operational roadmap toward physically-constrained Agricultural Digital Twins.

Land Surface Temperature (LST) plays a key role in climate monitoring, urban heat assessment, and land–atmosphere interactions. However, current thermal infrared satellite sensors cannot simultaneously achieve high spatial and temporal resolution. Spatio-temporal fusion (STF) techniques address this limitation by combining complementary satellite data, one with high spatial but low temporal resolution, and another with high temporal but low spatial resolution. Existing STF techniques, from classical models to modern deep learning (DL) architectures, were primarily developed for surface reflectance (SR). Their application to thermal data remains limited and often overlooks LST-specific spatial and temporal variability. This study provides a focused review of DL-based STF methods for LST. We present a formal mathematical definition of the thermal fusion task, propose a refined taxonomy of relevant DL methods, and analyze the modifications required when adapting SR-oriented models to LST. To support reproducibility and benchmarking, we introduce a new dataset comprising 51 Terra MODIS-Landsat LST pairs from 2013 to 2024, and evaluate representative models to explore their behavior on thermal data.

Accurate and timely diagnosis of plant diseases plays a crucial role in ensuring global food security, safeguarding human health, and supporting the economic well-being of farmers. Currently, there is a growing interest in developing reliable deep learning systems for identifying various plant species and their associated diseases. In this respect, combining multiple CNN is an effective approach for improving disease classification since each individual model has unique design principles that enable them to extract different levels and types of features from input images. The present work introduces a novel fuzzy aggregation-based CNN ensemble for tomato leaf disease classification. Initially, five CNN architectures, namely, VGG16, ResNet152V2, MobileNetV2, InceptionV3, and DenseNet201, are designed for classifying tomato leaf diseases. Based on their diverse statistical characteristics, selected models are then integrated into a CNN ensemble. To achieve this step, we introduce a new fuzzy Max-Sum integral, built upon a fuzzy convex combination operator. Performance assessment is conducted on two datasets including 10 and 6 tomato disease classes. Further experiments are conducted by merging these datasets to include a broader range of disease categories. The results demonstrate that the proposed CNN ensemble outperforms both individual models and the CNN ensemble based on the classical Sugeno fuzzy integral.

Weed species classification represents an important step for the development of automated targeting systems that allow the adoption of precision agriculture practices. To reduce costs and yield losses caused by their presence. The identification of weeds is a challenging problem due to their shared similarities with crop plants and the variability related to the differences in terms of their types. Along with the variations in relation to changes in field conditions. Moreover, to fully benefit from deep learning-based methods, large fully annotated datasets are needed. This requires time intensive and laborious process for data labeling, which represents a limitation in agricultural applications. Hence, for the aim of improving the utilization of the unlabeled data, regarding conditions of scarcity in terms of the labeled data available during the learning phase and provide robust and high classification performance. We propose a deep semi-supervised approach, that combines consistency regularization with similarity learning. Through our developed deep auto-encoder architecture, experiments realized on the DeepWeeds dataset and inference in noisy conditions demonstrated the effectiveness and robustness of our method in comparison to state-of-the-art fully supervised deep learning models. Furthermore, we carried out ablation studies for an extended analysis of our proposed joint learning strategy.

Self-supervised learning (SSL) has emerged as a powerful technique for learning visual representations. While recent SSL approaches achieve strong results in global image understanding, they are limited in capturing the structured representation in scenes. In this work, we propose a self-supervised approach that progressively builds structured visual representations by combining semantic grouping, instance level separation, and hierarchical structuring. Our approach, based on a novel ProtoScale module, captures visual elements across multiple spatial scales. Unlike common strategies like DINO that rely on random cropping and global embeddings, we preserve full scene context across augmented views to improve performance in dense prediction tasks. We validate our method on downstream object detection tasks using a combined subset of multiple datasets (COCO and UA-DETRAC). Experimental results show that our method learns object centric representations that enhance supervised object detection and outperform the state-of-the-art methods, even when trained with limited annotated data and fewer fine-tuning epochs.