Abstract
Precision Agriculture (PA) has emerged as an approach to optimize production, comprise different technology and principles focusing on how to improve agricultural production. Currently, one of the main foundations of PA is the use of artificial intelligence, through deep learning (DL) algorithms. By processing large volumes of complex data, DL enhances decision-making and boosts farming efficiency. However, these methods are hungry for annotated data, which contrasts with the scarce availability of annotated agricultural data and the costs of annotation. Self-supervised learning (SSL) has emerged as a solution to tackle the lack of annotated agricultural data. This study presents a review of the application of SSL methods to computer vision tasks in the agricultural context. The aim is to create a starting point for professionals and scientists who intend to apply these methods using agricultural data. The results of 33 studies found in the literature are discussed, highlighting their pros and cons. In most of the studies, SSL outperformed its supervised counterpart, using datasets from 4000 to 60,000 samples. Potential directions for improving future research are suggested.

Abstract
Inherited retinal diseases (IRDs) are genetic disorders affecting photoreceptors and the retinal pigment epithelium, leading to progressive vision loss. Retinitis pigmentosa (RP), the most common IRD, manifests as night blindness, peripheral vision loss, and eventually central vision decline. RP is genetically diverse and can be categorized into non-syndromic and syndromic. Advanced imaging technologies such as fundus autofluorescence (FAF) and spectral-domain optical coherence tomography (SD-OCT) facilitate diagnosing and managing these conditions. The integration of artificial intelligence in analyzing retinal images has shown promise in identifying genes associated with RP. This study used a dataset from Portuguese public hospitals, comprising 2798 FAF images labeled for syndromic and non-syndromic RP across 66 genes. Three pre-trained models, Inception-v3, ResNet-50, and VGG-19, were used to classify these images, obtaining an accuracy of over 80% in the training data and 54%, 56%, and 54% in the test data for all models. Data preprocessing included class balancing and boosting to address variability in gene representation. Model performance was evaluated using some main metrics. The findings demonstrate the effectiveness of deep learning in automatically classifying retinal images for different RP-associated genes, marking a significant advancement in the diagnostic capabilities of artificial intelligence and advanced imaging techniques in IRD.

Abstract
COVID-19 is an extremely contagious respiratory sickness instigated by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Common symptoms encompass fever, cough, fatigue, and breathing difficulties, often leading to hospitalization and fatalities in severe cases. CTCovid19 is a novel model tailored for COVID-19 detection, specifically honing in on a distinct deep learning structure, ResNet-50 trained with ImageNet serves as the foundational framework for our model. To enhance its capability to capture pertinent features related to COVID-19 patterns in Computed Tomography scans, the network underwent fine-tuning through layer adjustments and the addition of new ones. The model achieved accuracy rates that went from 97.0 % to 99.8 % across three widely recognized and documented datasets dedicated to COVID-19 detection. © 2024 The Authors

Abstract
The authors wish to make the following corrections to this manuscript [...]

Abstract
This paper presents a novel approach to improve the conversion of interferometric synthetic aperture radar (InSAR) ascending and descending orbit measurements into horizontal and vertical deformation components, explicitly considering SAR product characteristics (acquisition geometry, resolution, and positional accuracy). Conventional decomposition methods use square grids, inadequately addressing directional biases associated with satellite images characteristics, reducing measurement accuracy. It is proposed optimized alternative geometries - rectangle, hexagon, and double inverted isosceles trapezoid (diIT) - derived from theoretical analysis of scatterer influence areas for Sentinel-1 imagery and calibrated data from the European ground motion service (EGMS). Validation was conducted comparing results against global navigation satellite system (GNSS) ground-truth data. Accuracy was quantitatively evaluated using deformation velocity (DV) and average Euclidean distance (ED) metrics. Results demonstrated an average 25% improvement in DV detection over traditional square grids, with only minor trade-offs, such as lower scatterer density and sub-millimetric increases in error for hexagon and diIT geometries.

Abstract
Monitoring the impacts of selective logging in tropical forests remains challenging due to the reliance on labor intensive field surveys. This study relies on the use of pre- and post-logging airborne LiDAR data to provide a precise and scalable method for quantifying canopy disturbances, carried out within the Sustainable Management Plan for the Jamari National Forest in Rond & ocirc;nia. The analysis of the airborne LiDAR data revealed a significant increase in canopy gaps after logging (F= 63.5,p <0.001 ), with canopy gaps corresponding to an average increase of 3.9 +/- 0.4% relative to the total plot area due to logging activities. The mean canopy gap area per felled tree was 158.29 m(2) ( +/- 35.7). A strong positive correlation was found between canopy gaps that emerged after logging and the logged AGB (18.4 +/- 1.7Mg ha(-1) ). A significant reduction in mean canopy height was also observed, decreasing from 26.26 +/- 0.40 m before logging to 24.62 +/- 0.33 m after logging (F= 9.86,p= 0.005) . The mean canopy gap area shifted from 40.68 +/- 2.30 m(2) to 77.07 +/- 2.82 m(2). Furthermore, there was an increase of 14.6% in the total number of gaps. The average Gini coefficient was 0.50 +/- 0.02 before logging and 0.64 +/- 0.01 in the post-logging areas and the average total impact on the canopy was 16.6 +/- 1.5% of the selectively logged area. The results obtained using the proposed methodology were consistent with field observations, demonstrating high accuracy of LiDAR-detected impacts when compared with inventory and GNSS data. This high detection rate highlights the sensitivity of LiDAR point cloud data in capturing small structural changes. Compared to pre-logging conditions, the observed alterations demonstrate that LiDAR provides a more precise and scalable approach for quantifying the impact of selective logging on forest structure.