arXiv subjects

Francesco Petri

Publications and source records attributed to Francesco Petri.

5 recordsLinked to original sources

arXiv2026

An LLM-Based Automatic Sportscast Solution for Robot Soccer Matches

RoboCup has always been a scenario to develop systems that solve real-world problems. Driven by the main goal of playing against the 2050 FIFA World Cup champions, the RoboCup Soccer leagues need to constantly measure how the research community is progressing. Computing visual statistics from match videos is a crucial way to track this evolution. To address this challenge, this paper introduces a fully autonomous, real-time sports commentator for RoboCup matches. By bridging the gap between raw kinematic tracking and natural language generation, our neuro-symbolic architecture extracts precise statistics from video streams and turns them into fluent, hallucination-free narration. The proposed system is capable of generating statistics and commentary both during live match streaming and in post-game analysis, easily adapting to the new dynamism of the league where different humanoid robots of different sizes share the field. Supplemental materials are available at https://lab-rococo-sapienza.github.io/MARIO/

cs.CV↗

arXiv2025

Impact of redshift distribution uncertainties on Lyman-break galaxy cosmological parameter inference

A significant number of Lyman-break galaxies (LBGs) with redshifts 3 < z < 5 are expected to be observed by the upcoming Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST). This will enable us to probe the universe at higher redshifts than is currently possible with cosmological galaxy clustering and weak lensing surveys. However, accurate inference of cosmological parameters requires precise knowledge of the redshift distributions of selected galaxies, where the number of faint objects expected from LSST alone will make spectroscopic based methods of determining these distributions extremely challenging. To overcome this difficulty, it may be possible to leverage the information in the large volume of photometric data alone to precisely infer these distributions. This could be facilitated using forward models, where in this paper we use stellar population synthesis (SPS) to estimate uncertainties on LBG redshift distributions for a 10 year LSST (LSSTY10) survey. We characterise some of the modelling uncertainties inherent to SPS by introducing a flexible parameterisation of the galaxy population prior, informed by observations of the galaxy stellar mass function (GSMF) and cosmic star formation density (CSFRD). These uncertainties are subsequently marginalised over and propagated to cosmological constraints in a Fisher forecast. Assuming a known dust attenuation model for LBGs, we forecast constraints on the sigma8 parameter comparable to Planck cosmic microwave background (CMB) constraints.

astro-ph.CO↗

arXiv2025

Learning Local Causal World Models with State Space Models and Attention

World modelling, i.e. building a representation of the rules that govern the world so as to predict its evolution, is an essential ability for any agent interacting with the physical world. Despite their impressive performance, many solutions fail to learn a causal representation of the environment they are trying to model, which would be necessary to gain a deep enough understanding of the world to perform complex tasks. With this work, we aim to broaden the research in the intersection of causality theory and neural world modelling by assessing the potential for causal discovery of the State Space Model (SSM) architecture, which has been shown to have several advantages over the widespread Transformer. We show empirically that, compared to an equivalent Transformer, a SSM can model the dynamics of a simple environment and learn a causal model at the same time with equivalent or better performance, thus paving the way for further experiments that lean into the strength of SSMs and further enhance them with causal awareness.

cs.LG↗

arXiv2024

Real-Time Multimodal Signal Processing for HRI in RoboCup: Understanding a Human Referee

Advancing human-robot communication is crucial for autonomous systems operating in dynamic environments, where accurate real-time interpretation of human signals is essential. RoboCup provides a compelling scenario for testing these capabilities, requiring robots to understand referee gestures and whistle with minimal network reliance. Using the NAO robot platform, this study implements a two-stage pipeline for gesture recognition through keypoint extraction and classification, alongside continuous convolutional neural networks (CCNNs) for efficient whistle detection. The proposed approach enhances real-time human-robot interaction in a competitive setting like RoboCup, offering some tools to advance the development of autonomous systems capable of cooperating with humans.

cs.CV↗

arXiv2024

Transformers and Slot Encoding for Sample Efficient Physical World Modelling

World modelling, i.e. building a representation of the rules that govern the world so as to predict its evolution, is an essential ability for any agent interacting with the physical world. Recent applications of the Transformer architecture to the problem of world modelling from video input show notable improvements in sample efficiency. However, existing approaches tend to work only at the image level thus disregarding that the environment is composed of objects interacting with each other. In this paper, we propose an architecture combining Transformers for world modelling with the slot-attention paradigm, an approach for learning representations of objects appearing in a scene. We describe the resulting neural architecture and report experimental results showing an improvement over the existing solutions in terms of sample efficiency and a reduction of the variation of the performance over the training examples. The code for our architecture and experiments is available at https://github.com/torchipeppo/transformers-and-slot-encoding-for-wm

cs.LG↗