Accepted papers

We present a feasibility study for the use of a generative, probabilistic model, a Variational Autoencoder (VAE), to detect deviations from Standard Model (SM) physics in an electroweak process at the Large Hadron Collider (LHC). The new physics responsible for the anomalies is described through an Effective Field Theory (EFT) approach: the SM Lagrangian is Taylor-expanded and the higher order terms cause deviations in the kinematic distributions of the observables, and are thus identified by the model as anomalous contributions with respect to SM. Since the training of the model involves mainly only SM events, the proposed strategy is largely independent from any assumption on the nature of the new physics signature. To test the proposed strategy we use parton level generations of Vector Boson Scattering (VBS) events at the LHC, assuming an integrated luminosity of 350/fb.

Muon radiography is an advanced imaging technique that utilizes cosmic muons to visualize the interior of structures and materials, making it highly valuable for subsurface investigations. In this study, we present a measurement conducted using muon radiography at the Temperino mine. We demonstrate the application of an adaptive binning approach using Voronoi tessellation to enhance image visualization and improve cavity detection. The results reveal that the adaptive binning technique significantly improves the visibility of regions with cavities. The combination of muon radiography and adaptive binning through Voronoi tessellation showcases its potential as a powerful tool for subsurface exploration and geological studies, providing a more accurate and reliable approach for cavity detection and characterization.

In the rapidly evolving domain of fluorescence microscopy, the application of deep learning techniques for automatic cell segmentation presents exciting opportunities and challenges. In this work, we investigate the impact of loss functions and evaluation metrics on model performance and generalization in the context of cell recognition. First, we present extensive experiments with different commonly used loss functions and offer practical insights and guidelines, underscoring how the choice of a loss function can influence model performance. Second, we conduct a detailed examination of several evaluation metrics with their relative benefits and drawbacks, helping to guide effective model evaluation and comparison in the field. hird, we discuss how characteristics specific to fluorescence microscopy data impact model generalization. Precisely, we examine how factors such as cell sizes, color irregularities, and textures can potentially affect the performance and adaptability of these models to new data. Collectively, these insights provide a nuanced understanding of deep learning for automatic cell segmentation, shedding light on best practices, evaluation strategies, and model generalization. We hope this study can serve as a beneficial resource for researchers and practitioners working on similar applications, fostering further advancements in the field.

At LABEC (the INFN ion beam laboratory of nuclear techniques for environment and cultural heritage, located in Florence), a novel transportable accelerator for in-situ ion-beam analysis (IBA) measurements of cultural heritage materials, MACHINA, has been constructed as part of an international collaboration between the INFN and the CERN. Here we present the most recent developments regarding this project, consisting in the design, construction, and testing of the hardware/software of a data acquisition (DAQ) system for prompt photons of characteristic energies emitted following the interaction between charged particles and matter to construct elemental maps, i.e., grayscale dig ital images showing the spatial distribution of elements in a material of interest wherein the brightness of each individual pixel correlates to the abundance of a given element.

During the COVID-19 pandemic, a huge amount of literature was produced covering different aspects of infection. The use of artificial intelligence (AI) in medical imaging has been shown to improve screening, diagnosis, treatment, and medication for the COVID-19 virus. Applying natural language processing (NLP) solutions to COVID-19 literature has contributed to infer significant COVID-19-related topics and correlated diseases. In this paper, we aim at evaluating biomedical transformer-based NLP techniques in COVID-19 research to understand if they are able to classify problems related to COVID-19. Particularly, once collected COVID-19 publications encompassing the terms AI and medical imaging, fifteen BERT-based models have been compared with respect to modality prediction and task prediction.

To enhance the deep understanding of artists’mindset and technical practices the detailed characterization of paintings require the net of brushstrokes on the surface to be considered as informative. Laser profilometry is being coupled with a macro x-ray fluorescence (MA-XRF) device as a non-destructive, non-invasive, portable tool for such an investigation. A set of different brushstrokes was surveyed by testing the necessary sampling conditions and verifying the compatibility with the MA-XRF working conditions. Eight different surveyed sample brushstrokes proved to be clearly informative of a variety of features useful to characterize the gesturality of the artist’s hand.