Images available on online sharing platforms have a high probability of being modified, with additional global transformations such as compression, resizing or filtering covering the possible alteration. Such manipulations impose many constraints forgery detection algorithms. This article presents framework improving robustness for image detection. The most important step our is to take into account quality corresponding chosen application. Therefore, we relied camera identification model...

Optimizing deep learning models requires large amounts of annotated images, a process that is both time-intensive and costly. Especially for semantic segmentation in which every pixel must be annotated. A potential strategy to mitigate annotation effort active learning. Active facilitates the identification selection most informative images from unlabelled pool. The underlying premise these selected can improve model's performance faster than random reduce effort. While has demonstrated...

As deep learning predictive models become an integral part of a large spectrum precision agricultural systems, barrier to the adoption such automated solutions is lack user trust in these highly complex, opaque and uncertain models. Indeed, neural networks are not equipped with any explicit guarantees that can be used certify system's performance, especially varying uncontrolled environments as ones typically faced computer vision for agriculture.Fortunately, certain methods developed other...

The objectives of this study were to calculate the frequency hysterectomies at Conakry university hospitals (Donka Hospital and Ignace Deen Hospital), describe women's social, demographic, clinical characteristics, identify key indications, surgical techniques used, prognosis. This was a 2-year descriptive study, retrospective for first year (May 2011-April 2012) prospective second 2012-April 2013), 333 consecutive performed in obstetrics gynecology departments these two hospitals....

The objectives of this study were to calculate the frequency ectopic pregnancy in department, define its epidemiological, diagnostic, therapeutic, and prognostic aspects, determine a clear therapeutic approach appropriate our setting. In prospective study, we compiled all cases seen 2011 2012 obstetrics-gynecology department at Ignace Deen University Hospital Conakry. Ectopic pregnancies represented 1.3% deliveries over period. 111 population, women aged 30-34 years accounted for 31.5%,...

Objectives: To calculate the frequency of uterine fibromyomas, to describe epidemiological profile, management and establish prognosis women with fibromyomas in gynaecology-obstetrics department Regional Hospital Labé. Methodology: This was a retrospective descriptive study lasting three (3) years from 1 January 2017 31 December 2019. All patients admitted operated on for during period were included. Patients who not incomplete records Our data collected analysed using Epi-Info software...

The deployment of safe and trustworthy machine learning systems, particularly complex black box neural networks, in real-world applications requires reliable certified guarantees on their performance. conformal prediction framework offers such formal by transforming any point into a set predictor with valid, finite-set, the coverage true at chosen level confidence. Central to this methodology is notion nonconformity score function that assigns each example measure ''strangeness'' comparison...

This study assessed the impact of different compost doses (0, 20 and 30 t/ha) on growth parameters biomass yields two forage species (Brachiariaruziziensis Panicum maximum) in sandy-clay-silty soil tropical climates. Climate data showed favorable conditions 2022 compared to 2021, with an increase rainfall, relative humidity average temperatures. Soil physico-chemical analyzes a decrease organic matter nutrients control plots, confirming importance input maintain fertility. The results...

Introduction: Innovations in surgical and anaesthetic techniques to provide a good maternal child safety have made the cesarean section routine intervention obstetrics. However, its complications, especially short or long term per postoperative ones, are not exceptional can affect mothers’ vital prognosis. The objectives of this study were describe complications encountered during section. Patients method: This is prospective, descriptive analytical carried out at maternity ward Kankan...

A considerable number of metrics can be used to evaluate the performance machine learning algorithms. While much work is dedicated study and improvement data quality models’ performance, less research focused on these evaluation metrics, their intrinsic relationship, interplay influence among models, data, environments conditions in which they are applied. some works have been conducted general tasks such as classification, fewer efforts more complex problems object detection image...

Food and social safety nets have a history as long human civilization. In hunter gatherer societies, food sharing is pervasive. Group members who prove unlucky in the short run, hunting or foraging, receive from other households anticipation of reciprocal consideration at later time (Smith 1988). With emergence first large sedentary civilizations Middle East, administrative systems developed specifically around storage distribution. The ancient Egyptians, for example, stored public...

Objectives: The objectives of this work were to calculate the frequency obstetrical emergencies, describe socio-demographic profile women admitted for obstetric identify main care taking emergencies and establish maternal foetal prognosis emergencies. Methodology: It was a 6-month descriptive prospective study conducted in Obstetrics Gynaecology Department Donka National Hospital, CHU Conakry, Guinea. took place from July 1st December 31st, 2005. data collected entered corrected using Word...

Objectives: The objectives of this article are to: 1) Determine the prevalence parturient that escapes from ETME, 2) rate acceptability or non-acceptability HIV tests, 3) Describe epidemiological profile and 4) factors explaining lack screening during pregnancy. Methodology: This report is based on an eventual descriptive study carried out in Maternity Ward at Ratoma Medical Centre for past six months. took place over (6) months June 15 to December 15, 2016. data were collected by individual...

As deep learning predictive models become an integral part of a large spectrum precision agricultural systems, barrier to the adoption such automated solutions is lack user trust in these highly complex, opaque and uncertain models. Indeed, neural networks are not equipped with any explicit guarantees that can be used certify system's performance, especially varying uncontrolled environments as ones typically faced computer vision for agriculture.Fortunately, certain methods developed other...