Department of Computer Science

Sapienza, University of Rome

About us

Welcome to the GLADIA research lab! Based in the beautiful city of Rome, we are a dynamic team of computer scientists, physicists, engineers, and mathematicians who are passionate about AI and love to push the boundaries of what is possible, seeking innovative solutions to real-world problems – or sometimes, making up new ones. We work with images, molecules, graphs, 3D shapes, music, language, games, medical data, and quantum particles. We are always excited to collaborate with other disciplines, bring our expertise to the table and have fun with science!

Google Research Award "FairGeometry"
ERC StG 2018 "SPECGEO"

News

The 2nd edition of our UniReps Workshop on Unifying Representations in Neural Models has been accepted at NeurIPS 2024. Come join us!

Jul 18, 2024

Our paper "ReMatching: Low-Resolution Representations for Scalable Shape Correspondence" was accepted at ECCV 2024 main conference!

Jul 5, 2024

We have 2/2 papers accepted at ICASSP 2024!

Apr 14, 2024

Our paper "Vector Quantile Regression on Manifolds" was accepted at AISTATS 2024 main conference!

Jan 20, 2024

Our paper "Multi-Source Diffusion Models for Simultaneous Music Generation and Separation" was accepted as Oral (Top 1.2%) at ICLR 2024 main conference!

Jan 16, 2024

Featured Publications

Find the complete list in publications.

Multi-Source Diffusion Models for Simultaneous Music Generation and Separation

In this work, we define a diffusion-based generative model capable of both music synthesis and source separation by learning the score …

Giorgio Mariani, Irene Tallini, Emilian Postolache, Michele Mancusi, Luca Cosmo, Emanuele Rodolà

Multi-Source Diffusion Models for Simultaneous Music Generation and Separation

ASIF: Coupled Data Turns Unimodal Models to Multimodal Without Training

CLIP proved that aligning visual and language spaces is key to solving many vision tasks without explicit training, but required to …

Antonio Norelli, Marco Fumero, Valentino Maiorca, Luca Moschella, Emanuele Rodolà, Francesco Locatello

ASIF: Coupled Data Turns Unimodal Models to Multimodal Without Training

Accelerating Transformer Inference for Translation via Parallel Decoding

Autoregressive decoding limits the efficiency of transformers for Machine Translation (MT). The community proposed specific network …

Andrea Santilli, Silvio Severino, Emilian Postolache, Valentino Maiorca, Michele Mancusi, Riccardo Marin, Emanuele Rodolà

Relative representations enable zero-shot latent space communication

Neural networks embed the geometric structure of a data manifold lying in a high-dimensional space into latent representations. …

Luca Moschella, Valentino Maiorca, Marco Fumero, Antonio Norelli, Francesco Locatello, Emanuele Rodolà

Relative representations enable zero-shot latent space communication

Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning

Predicting interactions between proteins and other biomolecules solely based on structure remains a challenge in biology. A high-level …

Pablo Gainza, Freyr Sverrisson, Federico Monti, Emanuele Rodolà, Davide Boscaini, Michael M. Bronstein, Bruno Correia

Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning

LIMP: Learning Latent Shape Representations with Metric Preservation Priors

In this paper, we advocate the adoption of metric preservation as a powerful prior for learning latent representations of deformable 3D …

Luca Cosmo, Antonio Norelli, Oshri Halimi, Ron Kimmel, Emanuele Rodolà

Isospectralization, or how to hear shape, style, and correspondence

The question whether one can recover the shape of a geometric object from its Laplacian spectrum (‘hear the shape of the …

Luca Cosmo, Mikhail Panine, Arianna Rampini, Maks Ovsjanikov, Michael M. Bronstein, Emanuele Rodolà

Isospectralization, or how to hear shape, style, and correspondence

Deep Functional Maps: Structured Prediction for Dense Shape Correspondence

We introduce a new framework for learning dense correspondence between deformable 3D shapes. Existing learning based approaches model …

Or Litany, Tal Remez, Emanuele Rodolà, Alex M. Bronstein, Michael M. Bronstein

Deep Functional Maps: Structured Prediction for Dense Shape Correspondence

Projects

Open Source

An Italian instruction-tuned LLaMA

Andrea Santilli

Camoscio

Collection of manim components, layouts and slide templates aimed to ease the development of animations, with particular focus on live presentations.

Powermanim

Newton's Fractals on Surfaces

This repository contains a Unity3D project implementing a demo for the method presented in Newton’s Fractals on Surfaces via Bicomplex Algebra, published at ACM SIGGRAPH 2022 in the Poster program.

Filippo Maggioli

Newton's Fractals on Surfaces

Any Order FEM Laplacian

Finite Element Method (FEM) discretization of the Laplace-Beltrami operator on triangular meshes, parametric on the polynomial’s order of the hat functions.

Filippo Maggioli

Any Order FEM Laplacian

Popular template to bootstrap the scaffolding for your PyTorch project with PyTorch Lightning, Hydra, DVC, Weights and Biases, and Streamlit.

Valentino Maiorca, Luca Moschella

NN Template

Contact

rodola@di.uniroma1.it
Via Salaria, 113, Rome, RM 00198
Follow Us on GitHub!