Publications by authors named "F M Neri"
Cloud cover experiences rapid fluctuations, significantly impacting the irradiance reaching the ground and causing frequent variations in photovoltaic power output. Accurate detection of thin and fragmented clouds is crucial for reliable photovoltaic power generation forecasting. In this paper, we introduce a novel cloud detection method, termed Adaptive Laplacian Coordination Enhanced Cross-Feature U-Net (ALCU-Net).
View Article and Find Full Text PDFGenetically modified maize DP51291 was developed to confer control against susceptible corn rootworm pests and tolerance to glufosinate-containing herbicide; these properties were achieved by introducing the and expression cassettes. The molecular characterisation data and bioinformatic analyses do not identify issues requiring food/feed safety assessment. None of the identified differences in the agronomic/phenotypic and compositional characteristics tested between maize DP51291 and its conventional counterpart needs further assessment, except for phosphorus in forage and manganese, proline, oleic acid (C18:1) and linoleic acid (C18:2) in grain, which do not raise safety and nutritional concerns.
View Article and Find Full Text PDFCognitive fatigue (CF) increases accident risk reducing performance, especially during complex tasks such as driving. We evaluated whether transcranial random noise stimulation (tRNS) could mitigate CF and improve driving performance. In a double-blind study, thirty participants performed a virtual reality truck driving task during real ( = 15) or sham ( = 15) tRNS applied bilaterally on the "anti-fatigue network".
View Article and Find Full Text PDFArticle Synopsis
- Machine learning algorithms are used for counting people in crowds, and test-time adaptation methods help models adjust to specific conditions during testing by altering model parameters and using additional data augmentation.
- Unlike traditional methods that require extensive unannotated data for each new target domain, the proposed CrowdTTA approach combines test-time adaptation with meta-learning for better adaptability in unknown conditions.
- CrowdTTA generates pseudo labels using uncertainty from a dropout layer, facilitating dual-level optimization which improves model performance and accuracy in counting people under various crowd densities and scales during testing.