Arnaud Poletto

Computer Science Engineer

Hello, I'm Arnaud Poletto. I completed my Master's in Computer Science at EPFL, specializing in artificial intelligence and data science. I focus on applied machine learning and computer vision, developing deep learning and data-driven approaches across diverse domains, from understanding human perception in built environments to advancing healthcare technology.

Scroll down to see timeline

Timeline

My academic and professional journey

July 2025 - Now

Scientific Assistant at LTS5 & Undae Science

Laboratory of Signal Processing 5, EPFL

Current Focus
  • • Developing Python signal processing algorithms for contactless sleep monitoring
  • • Researching novel data-driven approaches for biomedical signal analysis
March 2025 - June 2025

Research Intern at LIPID

Laboratory of Integrated Performance in Design, EPFL

Key Achievements
  • • Developed a contrastive learning model and automated pipeline for sky characterization, extracting daylight metrics (cloud coverage, dynamics) from 21,000+ images
  • • Created a practical camera-based alternative to expensive sky-scanning equipment for architectural daylight research
September 2022 - February 2025

Master's Degree

M.Sc. Computer Science, EPFL

Final Grade 5.58/6
Program Details
  • • Specialization in artificial intelligence and data science
  • • Advanced coursework in machine learning and computer vision
  • • Research focus on data science applications

Temporal Human Visual Attention in Window Views

Master's Thesis

What do people actually look at when they gaze out a window? Using eye-tracking and a custom Graph Neural Network, I built a system that predicts visual attention in dynamic outdoor scenes, with applications for evidence-based architectural design.

Computer VisionDeep LearningSaliency PredictionDynamic Window ViewsVisual AttentionHuman Behavior
February 2024 - August 2024

Engineer Intern at Logitech

Logitech, Lausanne

Key Achievements
  • • Developed 6DoF tracking models for infrared stylus detection using PyTorch
  • • Built automated Blender pipeline generating 28,000+ annotated frames
  • • Designed multi-stage detection system combining temporal motion with spatial analysis
September 2023 - January 2024

LDR-HDR Luminosity Variability & Sky Dynamics

Research Publication

Can a regular camera replace expensive sky-scanning equipment for daylight research? I developed a deep learning pipeline to find out, analyzing 21,000+ sky images to characterize cloud dynamics and luminosity patterns.

Computer VisionDeep LearningDaylightSky LuminanceTemporal dynamicsLDR & HDR Imaging
March 2022 - August 2022

Software Engineer at Le Relais

Fondation Le Relais, Lausanne

Key Achievements
  • • Developed platform connecting teachers with 2,000+ students learning digital literacy
  • • Implemented file sharing, admin tools, and communication features using Laravel
  • • Managed full-stack development from backend to deployment
August 2021 - February 2022

Software Engineer at Unisanté

Unisanté, Lausanne

Key Achievements
  • • Extended admin platform and database infrastructure for medical system with 700,000+ patient records
  • • Built staff tools with role-based permissions and security features
  • • Implemented schema versioning, data migrations, and backward compatibility
September 2018 - July 2021

Bachelor's Degree

B.Sc. Computer Science, EPFL

Final Grade 4.92/6
Program Details
  • • Strong foundation in computer science fundamentals
  • • Core coursework in algorithms and data structures
  • • Programming languages and mathematical foundations