The availability of remote-sensing multisource data from optical-based satellite sensors has created new opportunities and challenges for forest monitoring in the Amazon Biome. In particular, change-detection analysis has emerged in recent decades to monitor forest-change dynamics, supporting some Brazilian governmental initiatives such as PRODES and DETER projects for biodiversity preservation in threatened areas. In recent years fully convolutional network architectures have witnessed numerous proposals adapted for the change-detection task. This paper comprehensively explores state-of-the-art fully convolutional networks such as U-Net, ResU-Net, SegNet, FC-DenseNet, and two DeepLabv3+ variants on monitoring deforestation in the Brazilian Amazon. The networks’ performance is evaluated experimentally in terms of Precision, Recall, $F1$-score, and computational load using satellite images with different spatial and spectral resolution: Landsat-8 and Sentinel-2. We also include the results of an unprecedented auditing process performed by senior specialists to visually evaluate each deforestation polygon derived from the network with the highest accuracy results for both satellites. This assessment allowed estimation of the accuracy of these networks simulating a process “in nature” and faithful to the PRODES methodology. We conclude that the high resolution of Sentinel-2 images improves the segmentation of deforestation polygons both quantitatively (in terms of $F1$-score) and qualitatively. Moreover, the study also points to the potential of the operational use of Deep Learning (DL) mapping as products to be consumed in PRODES.

Conditional (cGAN)-based methods proposed so far for synthetic aperture radar (SAR)-to-optical image synthesis tend to produce noisy and unsharp optical outcomes. In this work, we propose the atrous-cGAN, a novel cGAN architecture that improves the SAR-to-optical image translation. The proposed generator and discriminator networks rely on atrous convolutions and incorporate an atrous spatial pyramid pooling (ASPP) module to enhance fine details in the generated optical image by exploiting spatial context at multiple scales. This letter reports experiments carried out to assess the performance of atrous-cGAN for the synthesis of Landsat-8 images from Sentinel-1A data based on three public data sets. The experimental analysis indicated that the atrous-cGAN consistently outperformed the classical pix2pix counterpart in terms of visual quality, similar to the true optical image, and as a feature learning tool for semantic segmentation.

Many deep-learning (DL)-based, domain adaptation (DA) methods for remote sensing (RS) applications rely on adversarial training strategies to align features extracted from images of different domains in a shared latent space. However, the performance of such representation matching techniques is negatively impacted when class occurrences in the target domain, for which no labeled data are available during training, are highly imbalanced. In this work, we propose a DL-based representation matching approach for DA in the context of change detection tasks. We further evaluate the approach in a deforestation mapping application, characterized by a high-class imbalance between the deforestation and no-deforestation classes. The domains represent different sites in the Amazon and Brazilian Cerrado biomes. To mitigate the class imbalance problem, we devised an unsupervised pseudolabeling scheme based on change vector analysis (CVA) that prevents the feature alignment to be biased toward the overrepresented class. The experimental results indicate that the proposed approach can improve the accuracy of cross-domain deforestation detection.

This work proposes a new method for face and mouth tracking using video cameras. This method has been proposed to be used in intelligent spaces as an initial stage to provide information to higher level applications. The method incorporates the classifiers of the Viola and Jones detection method, modified to provide a probabilistic output previously proposed by the authors, useful for tracking with particle filter. The method combines classifiers trained to detect specific poses (frontal and side face views), building an independent likelihood model of pose changes. We also propose to combine the model in several cameras to allow the mouth tracking in a three-dimensional space. The system has been evaluated on the AV16.3 database sequences showing good results in both precision and recall when using a single camera (in a bidimensional space), and an error below 3cm when using three cameras (in a three-dimensional space).

Processing SAR images for sea ice mapping using deep learning is typically constrained to analyzing neighboring pixels rather than studying visual patterns at a large scale. This quality ignores the wide within-class variance and spatial non-stationary class statistics observed by covering complete scenes. To leverage global and local information and improve automated sea ice mapping, we introduce the Irregular Tokens on Transformers (ITT) method by uniquely combining convolutional neural networks (CNN), transformers, and unsupervised segmentation. The method is trained to capture long-range spatial dependencies to overcome non-stationary class statistics. In addition, ITT incorporates a novel loss function, including a pixel-based term that guides the semantic segmentation and a region-based term that promotes region-consistent feature maps, increasing the classifier confidence and improving the predictions along object boundaries. We evaluate the method by classifying ice and water on dual-polarized RADARSAT-2 scenes. The evaluation metrics depict the best performance, generally, when the training leverages both pixel and region level loss terms, after pre-training the CNN. In such a case, prediction maps are less sensitive to non-stationary patterns, more detailed on ice/water boundaries, and less uncertain separating the classes.

The growing availability of satellite observations and the scarcity of labeled data motivate the development of foundation models for automated sea ice analysis. Synthetic aperture radar (SAR), the primary data source for operational sea ice mapping, presents distinct challenges due to complex backscatter behavior and the reliance on coarse, polygon-based ice-chart annotations. In this work, in collaboration with the Canadian Ice Service, we introduce SAR-IceFM, an open-source SAR-based foundation model trained on Sentinel-1 data for multi-task Arctic sea ice analysis. Across a range of regimes for data and model scaling, SAR-IceFM consistently improves performance relative to training from scratch, with gains increasing for larger backbones and when sufficient supervised data is available for fine-tuning. The strongest improvements reach up to 6.6\% in a combined evaluation metric. We further provide a systematic empirical assessment of single-task versus multi-task fine-tuning, resolution transferability under single- and multi-scale training, and the impact of MAE mask ratio. Our results reveal task-dependent trade-offs, limitations of shared encoders under coarse supervision, and non-trivial interactions between resolution, supervision geometry, and generalization. Overall, the findings demonstrate that self-supervised pre-training is beneficial but not sufficient on its own: downstream performance remains strongly influenced by annotation quality, task heterogeneity, and resolution handling. To support reproducibility and future research, the SAR-IceFM codebase, experimental configurations, and sample models are publicly available at https://github.com/jnoat92/SAR-IceFM