Archivio della ricerca - Fondazione Bruno Kessler
Not a member yet
21227 research outputs found
Sort by
A search for dark matter produced in association with a dark Higgs boson decaying into a Higgs boson pair in 3b or 4b final states using pp collisions at √s = 13 TeV with the ATLAS detector
No full textA search is performed for dark matter particles produced in association with a resonant pair of Higgs bosons using 140 fb−1 of proton-proton collisions at a centre-of-mass energy of 13 TeV recorded by the ATLAS detector at the Large Hadron Collider. This signature is expected in some extensions of the Standard Model predicting the production of dark matter particles, and is interpreted in terms of a dark Higgs model containing a Z′ mediator in which the dark Higgs boson s decays into a pair of Higgs bosons. The dark Higgs boson is reconstructed through final states with at least three b-tagged jets, produced by the pair of Higgs boson decays, in events with significant missing transverse momentum consistent with the presence of dark matter. The observed data are found to be in good agreement with Standard Model predictions, constraining scenarios with dark Higgs boson masses within the range of 250 to 400 GeV and Z′ mediators up to 2.3 TeV
Deep Learning in Visual Odometry for Autonomous Driving
Get PDFPositioning, Navigation, and Timing (PNT) solutions are fundamental for autonomous driving, ensuring reliable localization for safe vehicle control in diverse environments. While GNSS-based systems provide absolute positioning, they become unreliable in GNSS-denied scenarios such as urban canyons or tunnels. Dead reckoning techniques, including Visual Odometry (VO), offer an alternative by estimating motion from onboard sensors. Integrating these methods with deep learning (DL) has shown potential for enhancing robustness, particularly in challenging conditions. This study, part of the VAIPOSA ESA project, investigates the performance of VO solutions under various environmental conditions using a simulation-based approach. The CARLA simulator provides controlled testing scenarios, enabling the evaluation of VO accuracy across different weather conditions, illumination changes, and dynamic environments. A synthetic stereo setup enables capturing error-free ground truth trajectories and fair evaluation of the VO methods. Multiple sequences are analyzed, reflecting real-world challenges such as poor visibility, texture variations, and occlusions. The findings highlight the influence of environmental factors and dynamic objects on VO performance and the role of DL in mitigating common failure modes
Measurement of double-differential charged-current Drell-Yan cross-sections at high transverse masses in pp collisions at √s= 13 TeV with the ATLAS detector
No full textThis paper presents a first measurement of the cross-section for the charged-current Drell-Yan process pp → W± → l±ν above the resonance region, where l is an electron or muon. The measurement is performed for transverse masses,
, between 200 GeV and 5000 GeV, using a sample of 140 fb−1 of pp collision data at a centre-of-mass energy of
= 13 TeV collected by the ATLAS detector at the LHC during 2015–2018. The data are presented single differentially in transverse mass and double differentially in transverse mass and absolute lepton pseudorapidity. A test of lepton flavour universality shows no significant deviations from the Standard Model. The electron and muon channel measurements are combined to achieve a total experimental precision of 3% at low
. The single- and double differential W-boson charge asymmetries are evaluated from the measurements. A comparison to next-to-next-to-leading-order perturbative QCD predictions using several recent parton distribution functions and including next-to-leading-order electroweak effects indicates the potential of the data to constrain parton distribution functions. The data are also used to constrain four fermion operators in the Standard Model Effective Field Theory formalism, in particular the lepton-quark operator Wilson coefficient
A Roman Carved Tale Modelled in 3D and Interpreted with AI
Get PDFThis study proposes an innovative methodology for documenting and semantically analysing cultural heritage by integrating artificial intelligence (AI) with a photogrammetric 3D model. The case study is the Trajan’s Column in Rome, a monumental structure adorned with a continuous helical relief depicting Emperor Trajan’s Dacian campaigns. AI-driven semantic segmentation is used to identify key elements (such as human figures, battle scenes and natural motifs) within the digitised sculptural narrative. Starting from a high-resolution photogrammetric 3D model, the column’s texture is divided into multiple segments and a multimodal large language model (MLLM) is applied to produce context-aware segmentation masks via natural language prompts. Results are then projected onto the 3D geometry and visualised through a web-based 3D viewer
Origami Fresnel Zone Plate Lens Reflector Antennas for Satellite Applications
No full textThis work presents a methodology for designing deployable reflector antennas that combine origami structures and the Fresnel zone plate lens to obtain a compact antenna structure. In particular, Miura and Yoshimura’s origami patterns have been considered for the design of the Fresnel reflector mirror and the conical horn antenna feeder, respectively. A set of memory-form alloy (MFA) actuators have been used to deploy the antenna. The MFA actuators are activated by a direct current aimed at increasing the temperature and activating the memorized shape. The combination of these techniques provides light, inexpensive, and very compact antennas, particularly suitable for satellite applications. A numerical and experimental assessment campaign has been carried out on antenna prototypes operating in the Ku band at 15 GHz. The obtained experimental results are quite promising
Promoting zirconia with carbon: enhanced hybrid ZrO2/C ketonization catalyst for the valorization of e-acetic acid
No full textTime Series Change Vector Analysis for Semisupervised Abrupt Land Cover Change Detection
No full textChange detection (CD) in satellite image time series (SITS) is more complex than in bitemporal images due to the higher dimensionality of the data. Utilizing the full dimensionality of the time series remains challenging, particularly with dense SITS. An approach that can minimize dimensions without compromising informational depth is essential. In this article, we present an innovative framework for change vector analysis (CVA) in time series analysis and initial demonstrations of its effectiveness in capturing the spectral–temporal characteristics of changes. Unlike current methods, the proposed approach incorporates a wide range of spectral–temporal information and constructs separate reference matrices for each change type, facilitating an in-depth analysis of change components for CD. Based on the time series change vector (TSCV), the proposed framework extends CVA into the time series perspective, offering novel interpretations for magnitude and direction across temporal and spectral dimensions. The framework’s effectiveness is validated using Sentinel-2 data, demonstrating significant improvements in tackling multiple CD challenges in dense SITS scenarios
Linear Transformers beat YOLO for Embedded Object Detection
No full textVision transformers (ViTs) have recently become the go-to standard for solving various computer vision tasks due to their superior performance and generalization capabilities. However, these architectures are complex to use in embedded and heavily resource-constrained devices for two main reasons: their high memory requirements and the use of complex operators seldom supported by embedded inference pipelines. Meanwhile, in embedded environments, it is still common to use older architectures with lower performance, but offering reduced memory consumption and higher compatibility with the limited embedded runtimes, usually supporting only a limited number of operators. In this paper, we present a neural architecture based on a novel linear transformer block capable of bridging the gap between the performance achieved by modern computer vision models and the broader support offered by architectures currently used in embedded environments. We also propose a solution for one-shot scaling of our architecture, called Hardware-Aware Scaling. This approach allows us to develop architectures tailored to embedded devices with different computational resources without requiring a lengthy network architecture search or manual architecture tuning. We tested our architecture on an object detection task and achieved performance comparable to recent versions of YOLO, with lower latency and parameter count while maximizing compatibility
2D and 3D Semantic Segmentation for Interpreting and Understanding 3D Heritage Spaces
Get PDFThe 3D digitization of Cultural Heritage (CH) sites has become increasingly requested for documentation, preservation, and analysis applications. Beyond capturing 3D spatial geometry, the semantic interpretation and understanding of digital models are critical for enabling meaningful CH studies and facilitating informed conservation strategies. However, manual annotation and classification of architectural elements and surface pathologies remain labor-intensive and time-consuming, underscoring the need for automated approaches. This study presents a comparative analysis between two distinct semantic segmentation frameworks: (1) a 2D-to-3D pipeline that projects 2D image-based detections onto 3D point clouds produced with V-SLAM data and (2) direct segmentation methods of 3D point clouds acquired with portable LiDAR sensors. These frameworks are evaluated on data acquired using two distinct mobile mapping systems (MMS): (1) a fisheye multi-camera Visual SLAM-based portable system (ATOM-ANT3D) for the 2D-to-3D pipeline; (2) a LiDAR-based MMS (Heron MS Twin Color) for the 3D segmentation methods. Achieved results demonstrate the ability of the proposed frameworks to generate semantically enriched 3D heritage data, with the 2D-to-3D method slightly outperforming the 3D segmentation techniques