Search

Malgré l’intérêt croissant pour la cartographie des fonds marins, seul un quart d’entre eux a pu être levé fidèlement à une résolution de l’ordre du km (30 arc-secondes). Ceci s’explique par les lourds coûts engendrés par les campagnes par bateau (sonar) et/ou par avion (lidar). Ainsi, la bathymétrie dérivée de satellite connaît un essor considérable depuis deux décennies. En fer de lance, l’imagerie multispectrale à très haute résolution spatiale de Pléiades Neo dispose de 6 bandes (4 visibles, 1 red edge, et 1 infrarouge) pourvues d’une résolution spatiale de 1,2 m, surclassant ainsi l’imagerie multispectrale de Pléiades-1 dotée de 4 bandes (3 visibles et 1 infrarouge) à 2 m. En s’appuyant sur un jeu de données Pléiades Neo 4 acquis au-dessus des eaux modérément turbides de la baie de Saint-Malo, ce travail a permis de quantifier les contributions des bandes deep blue et red edge à la prédiction de la bathymétrie lidar en regard de la nature de la modélisation, et de l’architecture du réseau neuronal. Premièrement, la modélisation semi-analytique (transfert radiatif) basée sur les transformées de ratio, testées individuellement et linéairement: le remplacement de la bande bleu par la bande deep blue a diminué la régression linéaire standard (bleu-vert-rouge-infrarouge, R2=0,36) de 11,1%, mais le remplacement de la bande rouge par red edge, a produit le même score. Deuxièmement, la modélisation semi-analytique basée sur la transformée de ratio Pléiades-1 versus la combinaison des 15 transformées Pléiades Neo : augmentations respectives de 52,8% de la régression linéaire (R2Pléiades-1=0,36, et R2Pléiades Neo=0,55), et de 36,2% de la régression non-linéaire par réseau de neurones à deux couches cachées à trois neurones (R2Pléiades-1=0,58, et R2Pléiades Neo=0,79). Finalement, la modélisation empirique basée sur la bonification spectrale de la simulation de Pléiades-1 par deep blue et red edge a généré un gain maximal de 6% pour la régression linéaire (R2Pléiades-1=0,50, et R2Pléiades Neo=0,53) et de 1,4% pour la régression non-linéaire par réseau de neurones à deux couches cachées à trois neurones (R2Pléiades-1=0,73, et R2Pléiades Neo=0,74).

The article details the developments and discussions within the Food Protection Trends Management Committee, emphasizing the significance of open access for articles and the upcoming Vice Chair election in 2025. Topics include enhancing content engagement through multimedia elements, recognizing outstanding reviewers, and exploring mentorship opportunities for students involved in food safety research.

Maintaining animals and also preserving ecology are great challenges because of the growing encroachment of people and various issues affecting the environment. In this perspective, a wide use of AI technology reveals the application with virtually limitless potential to transform the connotations of the conservation campaigns and to mitigate the challenges that all animals undergo globally. Therefore, this paper focuses on the various aspects AI in the establishment of a safer environment for animals in contact with people. Currently, there are so many phrased species gone extinct and the level of biodiversity decline increasing at a large scale. It is important to emphasize that the severity of this situation has been claimed by the Inter-governmental Platform on Biodiversity and Ecosystem Services Global Assessment Report on Biodiversity and Ecosystem Services. The main causes that correlate this tragedy are Habitat Degradation through human interferences, Poaching, Over-picketing of resources, Effecting the ecology through pollution, and Future effect of climate change. There are actions that are taken under various strategies like the current technology to enhance the protection of the species' divide. Technology has though helped out throughout the time in the preservation of the nature over the years. On this, it is quite clear that the examination and preservation of species has largely depended on the use of technology ranging from the simple radio-tracking methods to satellite images. Another method that enables the researchers to study with a certain degree of precision the displacements of the animals in question is radio telemetry. These implementing factors placed this study in a standby to reveal critical knowledge on the species' distribution, feeding habits, and mobility. Consequently, lessons of decision making on migration and areas of protection may be acquired by the conservationists. The integration of AI with the animal conservation within the recent years has brought amazing advances to our assessment, safety, as well as preservation of the earth's diverse species. Thus, the application of AI to the field of conservation is an applied science since it focuses on the analysis of big data as well as prediction. To put it simply, AI can be defined as the creation of software that may perform a plurality of elementary tasks with the immediate possibility of learning and drawing conclusions independently, based on the input data that is put to them. These algorithms engage with the exact skills to perceive patterns, identify items, and predict events in the future. Thus, contemporary society gets new opportunities to gather, analyze, and utilize data with the help of artificial intelligence in the questions of animal conservation. Artificial intelligence is that gel or glue or pulse that makes things happen, leads to more production. It empowers the researchers to control as well as evaluate information that was indeed extraordinarily cumbersome to control and evaluate before. It accelerates science, and therefore optimizes its efficiency, whether it is a case of distinguishing between numerous photos of cam traps to identify rare species or the analysis of big statistics to determine population density. This condition can turn increasing traditional ways of data harvesting into new forms that also considers data. It can contain information on current movements of animals, the environmental status as well as the threats which can contain aspects like poaching among others. In this chapter application of artificial intelligence in animal conservation will also be discussed; likewise, the factors influencing the effectiveness of wildlife conservation/reduction and control would also be reviewed; the limitations/ethical considerations regarding wildlife conservation and management would also be discussed. [ABSTRACT FROM AUTHOR]