Search Results for author: Matteo Matteucci

Found 60 papers, 15 papers with code

Latent Neural Cellular Automata for Resource-Efficient Image Restoration

no code implementations22 Mar 2024 Andrea Menta, Alberto Archetti, Matteo Matteucci

Neural cellular automata represent an evolution of the traditional cellular automata model, enhanced by the integration of a deep learning-based transition function.

Artificial Life Image Restoration

A Deep-Learning Technique to Locate Cryptographic Operations in Side-Channel Traces

1 code implementation29 Feb 2024 Giuseppe Chiari, Davide Galli, Francesco Lattari, Matteo Matteucci, Davide Zoni

Side-channel attacks allow extracting secret information from the execution of cryptographic primitives by correlating the partially known computed data and the measured side-channel signal.

Harnessing the Computing Continuum across Personalized Healthcare, Maintenance and Inspection, and Farming 4.0

no code implementations23 Feb 2024 Fatemeh Baghdadi, Davide Cirillo, Daniele Lezzi, Francesc Lordan, Fernando Vazquez, Eugenio Lomurno, Alberto Archetti, Danilo Ardagna, Matteo Matteucci

The AI-SPRINT project, launched in 2021 and funded by the European Commission, focuses on the development and implementation of AI applications across the computing continuum.

More than the Sum of Its Parts: Ensembling Backbone Networks for Few-Shot Segmentation

no code implementations9 Feb 2024 Nico Catalano, Alessandro Maranelli, Agnese Chiatti, Matteo Matteucci

\acrlong{fss}, in particular, concerns the extension and optimization of traditional segmentation methods in challenging conditions where limited training examples are available.

One-Shot Learning Segmentation +1

Can Shape-Infused Joint Embeddings Improve Image-Conditioned 3D Diffusion?

no code implementations2 Feb 2024 Cristian Sbrolli, Paolo Cudrano, Matteo Matteucci

We find that, while matching CLIP in generation quality and diversity, CISP substantially improves coherence with input images, underscoring the value of incorporating 3D knowledge into generative models.

Denoising Text-to-Image Generation

Deep Learning-based Target-To-User Association in Integrated Sensing and Communication Systems

no code implementations11 Jan 2024 Lorenzo Cazzella, Marouan Mizmizi, Dario Tagliaferri, Damiano Badini, Matteo Matteucci, Umberto Spagnolini

Simulation results over different urban vehicular mobility scenarios show that the proposed T2U method provides a probability of correct association that increases with the size of the BS antenna array, highlighting the respective increase of the separability of the VEs in the beamspace.

Age Group Discrimination via Free Handwriting Indicators

no code implementations29 Sep 2023 Eugenio Lomurno, Simone Toffoli, Davide Di Febbo, Matteo Matteucci, Francesca Lunardini, Simona Ferrante

Content-free handwriting data from 80 healthy participants in different age groups (20-40, 41-60, 61-70 and 70+) were analysed.

RadarLCD: Learnable Radar-based Loop Closure Detection Pipeline

no code implementations13 Sep 2023 Mirko Usuelli, Matteo Frosi, Paolo Cudrano, Simone Mentasti, Matteo Matteucci

The methodology undergoes evaluation across a variety of FMCW Radar dataset scenes, and it is compared to state-of-the-art systems such as Scan Context for Place Recognition and ICP for Loop Closure.

Image Retrieval Loop Closure Detection +2

Continual Cross-Dataset Adaptation in Road Surface Classification

no code implementations5 Sep 2023 Paolo Cudrano, Matteo Bellusci, Giuseppe Macino, Matteo Matteucci

Accurate road surface classification is crucial for autonomous vehicles (AVs) to optimize driving conditions, enhance safety, and enable advanced road mapping.

Autonomous Vehicles Classification +1

Scaling Survival Analysis in Healthcare with Federated Survival Forests: A Comparative Study on Heart Failure and Breast Cancer Genomics

no code implementations4 Aug 2023 Alberto Archetti, Francesca Ieva, Matteo Matteucci

Our results underscore the potential of FedSurF++ to improve the scalability and effectiveness of survival analysis in distributed settings while preserving user privacy.

Federated Learning Survival Analysis

Discriminative Adversarial Privacy: Balancing Accuracy and Membership Privacy in Neural Networks

no code implementations5 Jun 2023 Eugenio Lomurno, Alberto Archetti, Francesca Ausonio, Matteo Matteucci

The remarkable proliferation of deep learning across various industries has underscored the importance of data privacy and security in AI pipelines.

Bridging the Gap: Enhancing the Utility of Synthetic Data via Post-Processing Techniques

no code implementations17 May 2023 Andrea Lampis, Eugenio Lomurno, Matteo Matteucci

These results represent a new state of the art in Classification Accuracy Score and highlight the effectiveness of post-processing techniques in improving the quality of synthetic datasets.

Few Shot Semantic Segmentation: a review of methodologies and open challenges

no code implementations12 Apr 2023 Nico Catalano, Matteo Matteucci

Semantic segmentation assigns category labels to each pixel in an image, enabling breakthroughs in fields such as autonomous driving and robotics.

Autonomous Driving Few-Shot Semantic Segmentation +2

Two Steps Forward and One Behind: Rethinking Time Series Forecasting with Deep Learning

no code implementations10 Apr 2023 Riccardo Ughi, Eugenio Lomurno, Matteo Matteucci

The Transformer is a highly successful deep learning model that has revolutionised the world of artificial neural networks, first in natural language processing and later in computer vision.

Time Series Time Series Forecasting

A Multi-Modal Simulation Framework to Enable Digital Twin-based V2X Communications in Dynamic Environments

no code implementations13 Mar 2023 Lorenzo Cazzella, Francesco Linsalata, Maurizio Magarini, Matteo Matteucci, Umberto Spagnolini

Digital Twins (DTs) for physical wireless environments have been recently proposed as accurate virtual representations of the propagation environment that can enable multi-layer decisions at the physical communication equipment.

Federated Survival Forests

1 code implementation6 Feb 2023 Alberto Archetti, Matteo Matteucci

In this work, we present a novel federated algorithm for survival analysis based on one of the most successful survival models, the random survival forest.

Federated Learning Privacy Preserving +1

Enhancing Once-For-All: A Study on Parallel Blocks, Skip Connections and Early Exits

no code implementations3 Feb 2023 Simone Sarti, Eugenio Lomurno, Andrea Falanti, Matteo Matteucci

The use of Neural Architecture Search (NAS) techniques to automate the design of neural networks has become increasingly popular in recent years.

Knowledge Distillation Neural Architecture Search

Anticipate, Ensemble and Prune: Improving Convolutional Neural Networks via Aggregated Early Exits

no code implementations28 Jan 2023 Simone Sarti, Eugenio Lomurno, Matteo Matteucci

Today, artificial neural networks are the state of the art for solving a variety of complex tasks, especially in image classification.

Edge-computing Image Classification

IC3D: Image-Conditioned 3D Diffusion for Shape Generation

no code implementations20 Nov 2022 Cristian Sbrolli, Paolo Cudrano, Matteo Frosi, Matteo Matteucci

To address this limitation and enhance image-guided 3D DDPMs with augmented 3D understanding, we introduce CISP (Contrastive Image-Shape Pre-training), obtaining a well-structured image-shape joint embedding space.

3D Generation 3D Reconstruction +3

POPNASv2: An Efficient Multi-Objective Neural Architecture Search Technique

no code implementations6 Oct 2022 Andrea Falanti, Eugenio Lomurno, Stefano Samele, Danilo Ardagna, Matteo Matteucci

With its sequential model-based optimization strategy, Progressive Neural Architecture Search (PNAS) represents a possible step forward to face this resources issue.

Neural Architecture Search

On the utility and protection of optimization with differential privacy and classic regularization techniques

no code implementations7 Sep 2022 Eugenio Lomurno, Matteo Matteucci

According to the literature, this approach has proven to be a successful defence against several models' privacy attacks, but its downside is a substantial degradation of the models' performance.

L2 Regularization Privacy Preserving

Object Structural Points Representation for Graph-based Semantic Monocular Localization and Mapping

1 code implementation21 Jun 2022 Davide Tateo, Davide Antonio Cucci, Matteo Matteucci, Andrea Bonarini

In this paper, we propose the use of an efficient representation, based on structural points, for the geometry of objects to be used as landmarks in a monocular semantic SLAM system based on the pose-graph formulation.

Object Position +1

SGDE: Secure Generative Data Exchange for Cross-Silo Federated Learning

no code implementations24 Sep 2021 Eugenio Lomurno, Alberto Archetti, Lorenzo Cazzella, Stefano Samele, Leonardo Di Perna, Matteo Matteucci

In this context, federated learning is one of the most influential frameworks for privacy-preserving distributed machine learning, achieving astounding results in many natural language processing and computer vision tasks.

BIG-bench Machine Learning Fairness +2

Deep Learning of Transferable MIMO Channel Modes for 6G V2X Communications

no code implementations31 Aug 2021 Lorenzo Cazzella, Dario Tagliaferri, Marouan Mizmizi, Damiano Badini, Christian Mazzucco, Matteo Matteucci, Umberto Spagnolini

Algebraic Low-rank (LR) channel estimation exploits space-time channel sparsity through the computation of position-dependent MIMO channel eigenmodes leveraging recurrent training vehicle passages in the coverage cell.

Position Transfer Learning

Improving Multi-View Stereo via Super-Resolution

no code implementations28 Jul 2021 Eugenio Lomurno, Andrea Romanoni, Matteo Matteucci

Today, Multi-View Stereo techniques are able to reconstruct robust and detailed 3D models, especially when starting from high-resolution images.


Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*

2 code implementations4 May 2021 Alberto Archetti, Marco Cannici, Matteo Matteucci

Recently, the trend of incorporating differentiable algorithms into deep learning architectures arose in machine learning research, as the fusion of neural layers and algorithmic layers has been beneficial for handling combinatorial data, such as shortest paths on graphs.

Probabilistic electric load forecasting through Bayesian Mixture Density Networks

no code implementations23 Dec 2020 Alessandro Brusaferri, Matteo Matteucci, Stefano Spinelli, Andrea Vitali

Probabilistic load forecasting (PLF) is a key component in the extended tool-chain required for efficient management of smart energy grids.

Load Forecasting Management +1

Spatial Temporal Transformer Network for Skeleton-based Action Recognition

1 code implementation11 Dec 2020 Chiara Plizzari, Marco Cannici, Matteo Matteucci

Skeleton-based human action recognition has achieved a great interest in recent years, as skeleton data has been demonstrated to be robust to illumination changes, body scales, dynamic camera views, and complex background.

Action Recognition Skeleton Based Action Recognition +1

Facetwise Mesh Refinement for Multi-View Stereo

no code implementations1 Dec 2020 Andrea Romanoni, Matteo Matteucci

The refinement step is applied for each facet using only the camera pair selected.

3D Reconstruction

Estimation of Switched Markov Polynomial NARX models

no code implementations29 Sep 2020 Alessandro Brusaferri, Matteo Matteucci, Stefano Spinelli

This work targets the identification of a class of models for hybrid dynamical systems characterized by nonlinear autoregressive exogenous (NARX) components, with finite-dimensional polynomial expansions, and by a Markovian switching mechanism.


Identification of Probability weighted ARX models with arbitrary domains

no code implementations29 Sep 2020 Alessandro Brusaferri, Matteo Matteucci, Stefano Spinelli

Hybrid system identification is a key tool to achieve reliable models of Cyber-Physical Systems from data.

Skeleton-based Action Recognition via Spatial and Temporal Transformer Networks

1 code implementation17 Aug 2020 Chiara Plizzari, Marco Cannici, Matteo Matteucci

Skeleton-based Human Activity Recognition has achieved great interest in recent years as skeleton data has demonstrated being robust to illumination changes, body scales, dynamic camera views, and complex background.

Action Recognition In Videos Human Activity Recognition +1

A Differentiable Recurrent Surface for Asynchronous Event-Based Data

1 code implementation ECCV 2020 Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci

Dynamic Vision Sensors (DVSs) asynchronously stream events in correspondence of pixels subject to brightness changes.

Optical Flow Estimation

Towards Affordance Prediction with Vision via Task Oriented Grasp Quality Metrics

no code implementations10 Jul 2019 Luca Cavalli, Gianpaolo Di Pietro, Matteo Matteucci

Indeed, we assess the validity of our novel framework both in the context of perfect information, i. e., known object model, and in the partial information context, i. e., inferring task oriented metrics from vision, underlining advantages and limitations of both situations.


Mesh-based Camera Pairs Selection and Occlusion-Aware Masking for Mesh Refinement

no code implementations21 May 2019 Andrea Romanoni, Matteo Matteucci

Many Multi-View-Stereo algorithms extract a 3D mesh model of a scene, after fusing depth maps into a volumetric representation of the space.

TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo

no code implementations ICCV 2019 Andrea Romanoni, Matteo Matteucci

One of the most successful approaches in Multi-View Stereo estimates a depth map and a normal map for each view via PatchMatch-based optimization and fuses them into a consistent 3D points cloud.

A Data-driven Prior on Facet Orientation for Semantic Mesh Labeling

no code implementations26 Jul 2018 Andrea Romanoni, Matteo Matteucci

Mesh labeling is the key problem of classifying the facets of a 3D mesh with a label among a set of possible ones.

Attention Mechanisms for Object Recognition with Event-Based Cameras

no code implementations25 Jul 2018 Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci

Event-based cameras are neuromorphic sensors capable of efficiently encoding visual information in the form of sparse sequences of events.

Event-based vision Object Recognition +1

ReConvNet: Video Object Segmentation with Spatio-Temporal Features Modulation

no code implementations14 Jun 2018 Francesco Lattari, Marco Ciccone, Matteo Matteucci, Jonathan Masci, Francesco Visin

We introduce ReConvNet, a recurrent convolutional architecture for semi-supervised video object segmentation that is able to fast adapt its features to focus on any specific object of interest at inference time.

Object Position +3

Asynchronous Convolutional Networks for Object Detection in Neuromorphic Cameras

no code implementations21 May 2018 Marco Cannici, Marco Ciccone, Andrea Romanoni, Matteo Matteucci

Event-based cameras, also known as neuromorphic cameras, are bioinspired sensors able to perceive changes in the scene at high frequency with low power consumption.

Object object-detection +1

Predicting the Next Best View for 3D Mesh Refinement

1 code implementation16 May 2018 Luca Morreale, Andrea Romanoni, Matteo Matteucci

Finding the best poses to capture part of the scene is one of the most challenging topic that goes under the name of Next Best View.

3D Reconstruction Robot Navigation

Multi-View Stereo 3D Edge Reconstruction

1 code implementation17 Jan 2018 Andrea Bignoli, Andrea Romanoni, Matteo Matteucci

This paper presents a novel method for the reconstruction of 3D edges in multi-view stereo scenarios.

Mesh-based 3D Textured Urban Mapping

no code implementations18 Aug 2017 Andrea Romanoni, Daniele Fiorenti, Matteo Matteucci

In the era of autonomous driving, urban mapping represents a core step to let vehicles interact with the urban context.

Autonomous Driving Surface Reconstruction

Multi-View Stereo with Single-View Semantic Mesh Refinement

no code implementations16 Aug 2017 Andrea Romanoni, Marco Ciccone, Francesco Visin, Matteo Matteucci

In this paper we propose a novel method to refine both the geometry and the semantic labeling of a given mesh.

3D Reconstruction

Robust Moving Objects Detection in Lidar Data Exploiting Visual Cues

no code implementations29 Sep 2016 Gheorghii Postica, Andrea Romanoni, Matteo Matteucci

Detecting moving objects in dynamic scenes from sequences of lidar scans is an important task in object tracking, mapping, localization, and navigation.

Object Tracking

Sparse vs. Non-sparse: Which One Is Better for Practical Visual Tracking?

no code implementations30 Jul 2016 Yashar Deldjoo, Shengping Zhang, Bahman Zanj, Paolo Cremonesi, Matteo Matteucci

Recently, sparse representation based visual tracking methods have attracted increasing attention in the computer vision community.

Visual Tracking

Automatic 3D Reconstruction of Manifold Meshes via Delaunay Triangulation and Mesh Sweeping

no code implementations21 Apr 2016 Andrea Romanoni, Amaël Delaunoy, Marc Pollefeys, Matteo Matteucci

In this paper we propose a new approach to incrementally initialize a manifold surface for automatic 3D reconstruction from images.

3D Reconstruction

Incremental Reconstruction of Urban Environments by Edge-Points Delaunay Triangulation

no code implementations21 Apr 2016 Andrea Romanoni, Matteo Matteucci

Urban reconstruction from a video captured by a surveying vehicle constitutes a core module of automated mapping.

Efficient moving point handling for incremental 3D manifold reconstruction

no code implementations20 Jul 2015 Andrea Romanoni, Matteo Matteucci

From the 3D Delaunay triangulation of these points, state-of-the-art algorithms build a manifold rough model of the scene.

3D Reconstruction Management

Cannot find the paper you are looking for? You can Submit a new open access paper.