Hi there! My name is Gianluca Scarpellini. I'm a Research Scientist at Genesis Therapeutics working on generative models for structural biology, after a PhD in computer vision and RL at the Italian Institute of Technology and a research internship at Google DeepMind. I'm passionate about machine learning, research, and books!
Find me on
,
,
and
gianluca [at] scarpellini [dot] dev
.
Updates
-
2026-06
Preprint: New preprint from Genesis Therapeutics: Few-step Cofolding with All-Atom Flow Maps.
-
2026-01
Paper accepted!: Triangle Multiplication is All You Need for Biomolecular Structure Representations was accepted at ICLR2026!
-
2025-05
Preprint: New preprint: Pearl - A Foundation Model for Placing Every Atom in the Right Location.
-
2025-01
Paper published!: Condensation of Force Field Parameters from Machine Learning Predicted Distributions was published in the Journal of Chemical Information and Modeling.
-
2024-09
Paper accepted!: Re-assembling the past - The RePAIR dataset and benchmark for real world 2D and 3D puzzle solving was accepted at NeurIPS2024!
-
2024-07
Paper accepted!: My paper Look Around and Learn-Self-Training Object Detection by Exploration was accepted at ECCV2024!
-
2024-06
New Job: I started as research scientist at Genesis Therapeutics, working on generative models for structural biology.
-
2024-04
Paper accepted!: My paper DiffAssemble-A Unified Graph-Diffusion Model for 2D and 3D Reassembly was accepted at CVPR2024!
-
2024-01
Paper accepted!: My paper resulting from my internship at DeepMind π2vec: Policy Representation With Successor Features was accepted at ICLR2024!
-
2023-06
Paper published!: Positional diffusion: Ordering unordered sets with diffusion probabilistic models was published in Pattern Recognition Letters.
-
2022-09
DeepMind internship: I started an internship @ DeepMind, working on Offline Policy Evaluation and Successor Features
-
2022-01
Paper accepted @ WACV2022: My paper Event-driven re-id: A new benchmark and method towards privacy-preserving person re-identification was accepted at WACV2022
-
2021-05
1 Paper accepted @ CVPR2021: My paper on Monocular Human Pose Estimation with event-cameras was accepted at CVPRw on event-based vision
-
2020-11
PhD: I started a PhD on RL and computer vision at Istituto Italiano di Tecnologia
-
2020-09
Pytorch Lightning Contributor: I started contributing to the pytorch-lightning ecoysystem