Pier Luigi DovesiPhysical AI & AI for Social Good
Hi, I'm Pier — I lead The Good AI Lab, an independent AI lab working on foundational AI research and innovation projects for social good. Today we work with Doctors Without Borders (MSF), FMSI, and several universities around Europe.
By day, I lead the Robotics and Autonomous Systems team at AMD Silo AI, working on world action models, multimodal AI, and embodied intelligence.
My research spans domain adaptation, self-supervised learning, and real-time perception, with recent work extending toward vision-language models. Publications at NeurIPS, CVPR, ECCV, ICCV, 3DV, BMVC, and ICRA.
I also co-supervise 4 PhD researchers: Gabriele Pattarozzi and Mehdi Amlal at University of Bologna, and Nicola Ricciardi and Giovanni Grotto at University of Modena and Reggio Emilia.
Find me at GitHub, Google Scholar, LinkedIn, X, Hugging Face, ROCm Blog, and email.
Talks & Podcasts
AMD AI Research Club
(recurring)Peer-to-peer paper spotlight series · Host, 2025–present
info →NeurIPS 2026 — AI for Peace
(organizer)Paris, Dec 12–13, 2026
info →Bibliotheca Hertziana – Max Planck Institute
Research presentation · Nov 2, 2026
institute →ECCV 2026 — CV4GOOD
(host)Computer Vision for Humanitarian Action · Malmö, Sep 8, 2026
info →SYSTEC R&I Day 2026
(keynote)Faculty of Engineering, University of Porto · Jun 2026
info →ICLR 2026 — AI for Peace
(keynote)Rio de Janeiro, Apr 2026
info →GAIA Conference 2026
Where AI Finds Its Purpose — AI for humanitarian action with MSF · Gothenburg, Apr 2026
watch →Warsaw AI Summit 2026
(keynote)Warsaw, Mar 2026
event post →Compute to Impact
Advancing Physical AI · Helsinki, Mar 2026
info →Nordic HPC Summit 2025
(keynote)Where curiosity meets responsibility · Stockholm, Oct 2025
programme (PDF) →Redeye AI Theme Event
The Good AI Lab · Stockholm, Mar 2025
watch →AIAW Podcast E143
Revolutionizing AI and Autonomous Systems · Jan 2025
listen →Auto.AI Europe 2024
Data Alchemy — Berlin, Sep 2024
info →Nordic Industry Days 2024
Supercomputing for AI · Copenhagen, Sep 2024
info →GAIA Conference 2024
Embodied AI Beyond Datasets · Gothenburg, Apr 2024
info →News
Follow updates via RSS.
Presented my research at Bibliotheca Hertziana – Max Planck Institute (Nov 2)
Mind the Heads accepted at NeurIPS 26 — head-wise representation alignment (HeRA) for multimodal LLMs
Hosted AMD AI Research Club #5 — "Attack, Benchmark, Defend, Repeat: The Adversarial AI Cycle" with Antonio Cinà (University of Trieste)
Convened the World Model Foundry at the University of Trento, together with GBionics, the Universities of Bologna and Modena e Reggio Emilia
Workshop accepted at NeurIPS 2026 — AI for Peace (Paris, Dec 12–13, 2026)
Published "Efficient Hyperparameter Optimization for Autonomous Driving Models with AMD Instinct GPU Partitioning" on ROCm Blog
Hosted AMD AI Research Club #4 — "Rethinking Benchmarks for Robotic Manipulation" with Yi Ru (Helen) Wang (University of Washington)
Keynote at SYSTEC Research & Innovation Day (Faculty of Engineering, University of Porto, Jun 18)
Presented The Good AI Lab research at Sapienza University of Rome
Outstanding Reviewer at CVPR 2026 (top 5%)
Earlier updates (29)Hide earlier updates
Hosted AMD AI Research Club #3 — "Invisible Shortcuts: Why Vision Encoders Know Your Camera" with Giorgos Tolias (CTU in Prague)
Published "Semantic Fencing of Video Streams Using Embedding Splits from Vision Foundation Models" on ROCm Blog
Published "AMD-Powered 3D Gaussian Splatting for Autonomous Driving Scenes" on ROCm Blog
Published "Styled Text Image Generation with Eruku on AMD" on ROCm Blog
CV4GOOD accepted at ECCV 2026 — Computer Vision for Humanitarian Action workshop (Malmö, Sep 8, 2026), with MSF, featuring Yann Rebois, Jamon Van Den Hoek, Corey Scher, Joseph Redmon, and Sara Beery
Speaker at GAIA Conference 2026 (Gothenburg, Apr 20) — "Where AI Finds Its Purpose" — AI for humanitarian action with MSF
Keynote at ICLR 2026 — AI for Peace Workshop (Rio de Janeiro) — LAWS and international humanitarian law, alongside Timnit Gebru, Alex Hanna, David Gray Widder, Gisela Luján Andrade, and Joseph Redmon
Hosted AMD AI Research Club #2 — "Depth Foundation Models: Are We There Yet?" with Luca Bartolomei (University of Bologna)
Speaker at Compute to Impact (Helsinki) — "Advancing Physical AI: the AMD Open Ecosystem for Autonomy"
Launched AMD AI Research Club — "Retrieval Augmentation in Multimodal Large Language Models" with Sara Sarto (UniMoRE)
Keynote at Warsaw AI Summit 2026
Published "From Self-Driving Cars to Self-Driving Wars" — autonomy, AI weapons, and the illusion of control
"Seeing Beyond Words" — preprint on arXiv — JARVIS, a JEPA-inspired self-supervised visual learning framework for MLLMs
"From Principles to Practice" — our work with Doctors Without Borders
Published "3D Scene Reconstruction from the Inside: Explore the Mathematics Behind gsplat" on ROCm Blog
Keynote at Nordic HPC Summit 2025 (Stockholm) — "Building The Good AI Lab: where curiosity meets responsibility"
Published "Elevating 3D Scene Rendering with GSplat" on ROCm Blog
GAL and MSF announce partnership to bring responsible AI to humanitarian work
VocAlign accepted at BMVC 25 — source-free domain adaptation for open-vocabulary segmentation
SemLA accepted at CVPR 25 — domain adaptation with LoRA retrieval and fusion
Presented GAL at Redeye AI Theme Event 2025 (Stockholm)
Featured on AIAW Podcast E143 — Revolutionizing AI and Autonomous Systems
Speaker at Nordic Industry Days 2024 (Copenhagen) — Supercomputing for AI, EuroHPC
Speaker at Auto.AI Europe (Berlin) — "Data Alchemy: Adapting and Creating with Limited Data"
Speaker at GAIA Conference 2024 (Gothenburg) — "Adapting Like Humans: Embodied AI Beyond Datasets and Domains"
HAMLET accepted at ICCV 2023 — real-time online domain adaptation
OnDA accepted at ECCV 2022 — online domain adaptation
Real-Time Semantic Stereo Matching accepted at ICRA 2020
Enhancing Self-Supervised Monocular Depth Estimation with Traditional Visual Odometry accepted at 3DV 2019
Timeline
2025 – Present
The Good AI Lab — Co-Founder & Chair
Independent AI lab working on foundational AI research and innovation projects for social good.
2024 – Present
AMD Silo AI — Robotics Lead
part of AMD
Robotics team manager. Open ecosystem for Physical AI. World models, VLMs, and VLAs for humanoids and autonomous systems. Host of the AMD AI Research Club.
2023 – 2024
Silo AI — Senior AI Scientist & Expert Lead
LiDAR perception, LLMs, diffusion models.
2019 – 2023
Univrses — Deep Learning Engineer
consulting at Zenseact
Real-time perception and safety for autonomous driving. Lots of HPOs. Domain Adaptation research.
2017 – 2019
KTH Royal Institute of Technology — MSc Systems, Control & Robotics
2013 – 2017
University of Bologna & Tongji University — BSc Automation Engineering
Writing
- From Self-Driving Cars to Self-Driving Wars — autonomy, AI weapons, and the illusion of control (Jan 2026)
- From Principles to Practice — our work with Doctors Without Borders (Dec 2025)
- GAL and MSF partner to bring responsible AI to humanitarian work (Oct 2025)
soul.md
I think it's important now to tie care to Quality by pointing out that care and Quality are internal and external aspects of the same thing. A person who sees Quality and feels it as he works is a person who cares. A person who cares about what he sees and does is a person who's bound to have some characteristics of Quality.
— Robert M. Pirsig
Pessimismo della realtà, ottimismo della speranza.
— Antonio Gramsci
L'inferno dei viventi non è qualcosa che sarà; se ce n'è uno, è quello che è già qui, l'inferno che abitiamo tutti i giorni … Due modi ci sono per non soffrirne. Il primo riesce facile a molti: accettare l'inferno e diventarne parte fino al punto di non vederlo più. Il secondo è rischioso ed esige attenzione e apprendimento continui: cercare e saper riconoscere chi e cosa, in mezzo all'inferno, non è inferno, e farlo durare, e dargli spazio.
— Italo Calvino
Innovation without a guiding purpose is hollow. Real progress means more than building better tools; it means building them for everyone, especially those too often overlooked or forgotten. [...] We see AI as a force that reveals new paths, includes more voices, and helps reshape society for the better. If we are to contribute with a verse to this powerful play, let it be driven by curiosity and guided by responsibility.