Rosario Di Carlo

Rosario Di Carlo

Developer

Contact Me

About Me

I'm Rosario (Rs).
I have been passionate about programming since childhood, and I am constantly learning and growing in this field. As as a Computer Engineer my core expertise is developed in specialized areas: Computer Vision, Machine Learning, Big Data, and High-Performance Computing (HPC).
I always favor a hands-on, practical approach to solve problems, providing effective solutions.

Work Experience

Data Scientist - NATO CMRE (2024 - Present)

Scientist at NATO Centre for maritime research & experimentation.

  • Research on data fusion ML-based methods for maritime surveillance.
  • Apply on-board processing in underwater enviroments (Edge-AI, low power, TPU devices).

Research Engineer - Leonardo Labs (2021 - 2024)

Researcher at Leonardo's R&D laboratories.

  • Porting from IDL to Python of Datamet, sw for data acquisition and processing of Italian weather radar network.
  • Developing and researching Hyperspectral Image Analysis algorithms.
  • Designing and developing AI-based sensor fusion (AIS - Radar) service for the EU project PROMENADE.
  • Architecting the new LHD data infrastructure and migrating the old one (Spark, Hadoop, HDFS, Openstack)
  • Support HPC activities, deploy and containerizzation of AI-based applications
  • Teaching AI and Big Data at Italian workshops.

Research Fellow - AImageLab (2019 - 2021)

Researcher at AImageLab laboratory

  • Design and Development a system for operator-machinery interaction. (Tracking, 3D pose estimation, REBA score computing)

Timeline projects

Some of the projects here are based on material from my work that can be publicly disclosed, while others are created to develop new skills or just for fun in my free time and not with usability in mind.

04/01/2026

InkyPi plugin for e-paper displays

This plugin, for InkyPi, displays ambient sensor data from a temperature and humidity sensor (such as a BME680) and an inspirational Zen quote.

Find out more

28/05/2025
A model-based transformer approach for multitarget state estimation with a large number of sensors

[Article] A model-based transformer approach for multitarget state estimation with a large number of sensors

Target detection in a scenario with large number of sensors, how to mitigate performance degradation.

Find out more

19/08/2023
Improve target detection on hyperspectral aerial images using simulated atmosphere conditions

[Article] Improve target detection on hyperspectral aerial images using simulated atmosphere conditions

Data augmentation technique to improve target detection using simulated atmosphere conditions

Find out more

20/07/2023
ELSA project

ELSA

🛠️ Python (PyTorch)

These datasets have been generated using diffusion models in the context of the EU project ELSA for the Multimedia use-case. The goal is to develop effective solutions for detecting and mitigating the spread of deep fake images in multimedia content.

Datasets released on Hugginface:

21/05/2023

3D Metaverse

🛠️ Unity, C#

This project is a demonstration of the 3D Google tiles API and RealMe Avatars API. It allows users to explore real-world locations in a 3D digital environment and customize their avatars for an immersive experience.

Find out more

01/03/2023
Promenade vessel sensor fusion

PROMENADE H2020

🛠️ Backend: Python

The developed service fuses Automatic Identification System (AIS) and Radar messages using Machine Learning (ML) algorithms. The service is dockerized making it easy to deploy. The ML algorithms used enable accurate tracking and identification of maritime vessels, thereby enhancing maritime safety and security.

Find out more

05/03/2022
Distributed workflows with Jupyter

[Article] Distributed workflows with Jupyter

Multi-step simulation workflows in Quantum ESPRESSO in order to assess the Jupyter-workflow capabilities to enable interactive simulations of realistic, large-scale systems.

Find out more

01/03/2022
Hyperspectral image analysis

HyperHound

🛠️ Frontend: Python (Tkinter)
🛠️ Backend: Python (Pytorch)

HyperHound is a framework developed for hyperspectral imagery analysis. It implements different deep learning and classical models for target detection.

13/10/2021
EuroCC course

[Slides] EuroCC course

Advanced AI for industrial applications: Introduction to Compute Vision (Object detector)

Find out more

03/12/2020
Sequence modelling

[Slides] Sequence Modelling

State-of-the-art Deep learning models to handle sequential data.

Find out more

08/09/2020
Human machinery interaction

Human Machinery interaction

🛠️ Frontend: C#
🛠️ Backend: Python

The developed service is an advanced analytics tool that enables the analysis of the interaction between operators and machinery. This tool utilizes several deep learning models for detection and pose estimation, enabling it to accurately track operators movements and identify potentially dangerous actions. Additionally, it computes statistics on the safety of the working area.

Find out more

20/04/2020
Ergonomics metrics

Quarantine Notes

Android / iOS app built using Flutter framework to share notes during covid-19 quarantine

Find out more

19/12/2019
REBA score estimation

REBA score estimation

🛠️ Backend: Python

REBA Score Estimation is a Python package that enables the computation of ergonomics metrics, such as Rapid Entire Body Assessment (REBA) and Ovako Working Posture Analysis System (OWAS), from a human pose.

Find out more

04/03/2019
Natural Language Understading

Natural language Understading

Deep Learning toolkit for Italian Natural Language Understanding

Find out more

27/12/2018
Deep Entity Resolution

(Deep) Entity Resolution

Entity Resolution using deep learning

Find out more

21/10/2018
Visual Reasoning

Visual Reasoning

Module memory network and baselines models for visual reasoning in Pytorch on Clevr dataset

Find out more

29/04/2018
KNN classifier

Neural turing machine

Neural turing machine implementation in Pytorch

Find out more

01/01/2018
Big Gorilla - Open Refine

[Doc] BigGorilla vs OpenRefine

Data integration: Comparison between Big Gorilla and Open Refine

Find out more

17/11/2017
Rs-Crawler

Rs-Crawler

Multi-thread Web crawler in .Net, allows you to crawl websites’ URLs and fetch structure and pages information.

Find out more

31/01/2017
MindWave

MindWave

Measurement Electroencephalogram (EEG) activity through the device NeuroSky MindWave , detects wave's type and analyzes data patterns with neural networks.

Find out more

10/10/2016
SmartRoad

SmartRoad

“A journey is best measured in friends, rather than miles.” Leave Your Message in any place

Find out more

25/09/2015
project name

ImageCaption

Captioning photo on Android devices. Test with Deep Neural Network image recognition.

Find out more

09/09/2011
fb-friendsconnections

Facebook friends graph

View hierarchically your Facebook contacts depending on interactions.

Find out more

12/07/2011
RsWpaHack

RsWpaHack

App for Windows mobile 6.1 to test the security of Alice and Fastweb networks.

Find out more

10/05/2011
WriteProcessMemory

WriteProcessMemory tool

Simple tool for reading and writing memory in x86 and x64 processes.

Find out more

05/06/2010
RsWpaHack

FindOffset tool

Tool for reverse engineering. Calculate the distance between two instructions. Usefull with IDA pro.

Find out more