arXiv subjects

Francesco Verdoja

Publications and source records attributed to Francesco Verdoja.

At least 19 recordsLinked to original sources

arXiv2026

ReGIL: Retrieval-Guided Imitation Learning from a Single Demonstration

Learning robot manipulation policies with deep neural networks from a single demonstration remains highly challenging, as even small deviations from the demonstrated trajectory can quickly compound into failure, while collecting substantial online interaction data is costly. We propose ReGIL, a retrieval-guided imitation learning framework that treats a single demonstration as an external memory. ReGIL repeatedly queries this static memory throughout training to simultaneously guide exploration, generate the regularization buffer, and construct rewards. Specifically, it computes rewards through local temporal alignment between the current trajectory and the retrieved segment, providing step-wise and informative feedback for policy improvement. We evaluate ReGIL on robotic manipulation tasks from the LIBERO and Meta-World benchmarks under the single demonstration setting. ReGIL outperforms prior baselines in both success rate and training efficiency. In real-robot experiments, using only one demonstration and less than one hour of online training, ReGIL achieves over 75% success rate across three manipulation tasks with randomness in both initial robot pose and target position. These results demonstrate that leveraging the single demonstration as reusable memory can provide more than static supervision for efficient robot learning. More details can be found on our website: https://regil2026.github.io/

cs.RO↗

arXiv2026

Rheos: Modelling Continuous Motion Dynamics in Hierarchical 3D Scene Graphs

3D Scene Graphs (3DSGs) provide hierarchical, multi-resolution abstractions that encode the geometric and semantic structure of an environment, yet their treatment of dynamics remains limited to tracking individual agents. Maps of Dynamics (MoDs) complement this by modeling aggregate motion patterns, but rely on uniform grid discretizations that lack semantic grounding and scale poorly. We present Rheos, a framework that explicitly embeds continuous directional motion models into an additional dynamics layer of a hierarchical 3DSG that enhances the navigational properties of the graph. Each dynamics node maintains a semi-wrapped Gaussian mixture model that captures multimodal directional flow as a principled probability distribution with explicit uncertainty, replacing the discrete histograms used in prior work. To enable online operation, Rheos employs reservoir sampling for bounded-memory observation buffers, parallel per-cell model updates and a principled Bayesian Information Criterion (BIC) sweep that selects the optimal number of mixture components, reducing per-update initialization cost from quadratic to linear in the number of samples. Evaluated across four spatial resolutions in a simulated pedestrian environment, Rheos consistently outperforms the discrete baseline under continuous as well as unfavorable discrete metrics. We release our implementation as open source.

cs.RO↗

arXiv2026

Relational Scene Graphs for Object Grounding of Natural Language Commands

Robots are finding wider adoption in human environments, increasing the need for natural human-robot interaction. However, understanding a natural language command requires the robot to infer the intended task and how to decompose it into executable actions, and to ground those actions in the robot's knowledge of the environment, including relevant objects, agents, and locations. This challenge can be addressed by combining the capabilities of large language models (LLMs) to understand natural language with 3D scene graphs (3DSGs) for grounding inferred actions in a semantic representation of the environment. However, many 3DSGs lack explicit spatial relations between objects, even though humans often rely on these relations to describe an environment. This paper investigates whether incorporating open- or closed-vocabulary spatial relations into 3DSGs can improve the ability of LLMs to interpret natural language commands. To address this, we implement two pipelines using off-the-shelf models: an LLM-based pipeline for target object grounding from open-vocabulary language commands and a vision language model-based pipeline to add open-vocabulary spatial edges to 3DSGs from images captured while mapping. Finally, we evaluate two LLMs across 14 scenes using 905 natural language statements (786 procedurally-generated, 119 human-authored) to assess performance on the downstream task of target object grounding. Our study demonstrates that explicit spatial relations improve the ability of LLMs to ground objects, and while open-vocabulary relation generation with vision language models proves feasible from robot-captured images, our analysis did not yield evidence favoring either open- or closed-vocabulary relations.

cs.RO↗

arXiv2025

Event-Grounding Graph: Unified Spatio-Temporal Scene Graph from Robotic Observations

A fundamental aspect for building intelligent autonomous robots that can assist humans in their daily lives is the construction of rich environmental representations. While advances in semantic scene representations have enriched robotic scene understanding, current approaches lack a connection between spatial features and dynamic events; e.g., connecting the blue mug to the event washing a mug. In this work, we introduce the event-grounding graph (EGG), a framework grounding event interactions to spatial features of a scene. This representation allows robots to perceive, reason, and respond to complex spatio-temporal queries. Experiments using real robotic data demonstrate EGG's capability to retrieve relevant information and respond accurately to human inquiries concerning the environment and events within. Furthermore, the EGG framework's source code and evaluation dataset are released as open-source at: https://github.com/aalto-intelligent-robotics/EGG.

cs.RO↗

arXiv2025

QuASH: Using Natural-Language Heuristics to Query Visual-Language Robotic Maps

Embeddings from Visual-Language Models are increasingly utilized to represent semantics in robotic maps, offering an open-vocabulary scene understanding that surpasses traditional, limited labels. Embeddings enable on-demand querying by comparing embedded user text prompts to map embeddings via a similarity metric. The key challenge in performing the task indicated in a query is that the robot must determine the parts of the environment relevant to the query. This paper proposes a solution to this challenge. We leverage natural-language synonyms and antonyms associated with the query within the embedding space, applying heuristics to estimate the language space relevant to the query, and use that to train a classifier to partition the environment into matches and non-matches. We evaluate our method through extensive experiments, querying both maps and standard image benchmarks. The results demonstrate increased queryability of maps and images. Our querying technique is agnostic to the representation and encoder used, and requires limited training.

cs.RO↗

arXiv2025

MoDeSuite: Robot Learning Task Suite for Benchmarking Mobile Manipulation with Deformable Objects

Mobile manipulation is a critical capability for robots operating in diverse, real-world environments. However, manipulating deformable objects and materials remains a major challenge for existing robot learning algorithms. While various benchmarks have been proposed to evaluate manipulation strategies with rigid objects, there is still a notable lack of standardized benchmarks that address mobile manipulation tasks involving deformable objects. To address this gap, we introduce MoDeSuite, the first Mobile Manipulation Deformable Object task suite, designed specifically for robot learning. MoDeSuite consists of eight distinct mobile manipulation tasks covering both elastic objects and deformable objects, each presenting a unique challenge inspired by real-world robot applications. Success in these tasks requires effective collaboration between the robot's base and manipulator, as well as the ability to exploit the deformability of the objects. To evaluate and demonstrate the use of the proposed benchmark, we train two state-of-the-art reinforcement learning algorithms and two imitation learning algorithms, highlighting the difficulties encountered and showing their performance in simulation. Furthermore, we demonstrate the practical relevance of the suite by deploying the trained policies directly into the real world with the Spot robot, showcasing the potential for sim-to-real transfer. We expect that MoDeSuite will open a novel research domain in mobile manipulation involving deformable objects. Find more details, code, and videos at https://sites.google.com/view/modesuite/home.

cs.RO↗

arXiv2025

Minimal Intervention Shared Control with Guaranteed Safety under Non-Convex Constraints

Shared control combines human intention with autonomous decision-making. At the low level, the primary goal is to maintain safety regardless of the user's input to the system. However, existing shared control methods-based on, e.g., Model Predictive Control, Control Barrier Functions, or learning-based control-often face challenges with feasibility, scalability, and mixed constraints. To address these challenges, we propose a Constraint-Aware Assistive Controller that computes control actions online while ensuring recursive feasibility, strict constraint satisfaction, and minimal deviation from the user's intent. It also accommodates a structured class of non-convex constraints common in real-world settings. We leverage Robust Controlled Invariant Sets for recursive feasibility and a Mixed-Integer Quadratic Programming formulation to handle non-convex constraints. We validate the approach through a large-scale user study with 66 participants-one of the most extensive in shared control research-using a simulated environment to assess task load, trust, and perceived control, in addition to performance. The results show consistent improvements across all these aspects without compromising safety and user intent. Additionally, a real-world experiment on a robotic manipulator demonstrates the framework's applicability under bounded disturbances, ensuring safety and collision-free operation.

cs.RO↗

arXiv2025

REACT: Real-time Efficient Attribute Clustering and Transfer for Updatable 3D Scene Graph

Modern-day autonomous robots need high-level map representations to perform sophisticated tasks. Recently, 3D scene graphs (3DSGs) have emerged as a promising alternative to traditional grid maps, blending efficient memory use and rich feature representation. However, most efforts to apply them have been limited to static worlds. This work introduces REACT, a framework that efficiently performs real-time attribute clustering and transfer to relocalize object nodes in a 3DSG. REACT employs a novel method for comparing object instances using an embedding model trained on triplet loss, facilitating instance clustering and matching. Experimental results demonstrate that REACT is able to relocalize objects while maintaining computational efficiency. The REACT framework's source code will be available as an open-source project, promoting further advancements in reusable and updatable 3DSGs.

cs.RO↗

arXiv2024

Jointly Learning Cost and Constraints from Demonstrations for Safe Trajectory Generation

Learning from Demonstration allows robots to mimic human actions. However, these methods do not model constraints crucial to ensure safety of the learned skill. Moreover, even when explicitly modelling constraints, they rely on the assumption of a known cost function, which limits their practical usability for task with unknown cost. In this work we propose a two-step optimization process that allow to estimate cost and constraints by decoupling the learning of cost functions from the identification of unknown constraints within the demonstrated trajectories. Initially, we identify the cost function by isolating the effect of constraints on parts of the demonstrations. Subsequently, a constraint leaning method is used to identify the unknown constraints. Our approach is validated both on simulated trajectories and a real robotic manipulation task. Our experiments show the impact that incorrect cost estimation has on the learned constraints and illustrate how the proposed method is able to infer unknown constraints, such as obstacles, from demonstrated trajectories without any initial knowledge of the cost.

cs.RO↗

arXiv2024

Localization Under Consistent Assumptions Over Dynamics

Accurate maps are a prerequisite for virtually all mobile robot tasks. Most state-of-the-art maps assume a static world; therefore, dynamic objects are filtered out of the measurements. However, this division ignores movable but non-moving -- i.e., semi-static -- objects, which are usually recorded in the map and treated as static objects, violating the static world assumption and causing errors in the localization. This paper presents a method for consistently modeling moving and movable objects to match the map and measurements. This reduces the error resulting from inconsistent categorization and treatment of non-static measurements. A semantic segmentation network is used to categorize the measurements into static and semi-static classes, and a background subtraction filter is used to remove dynamic measurements. Finally, we show that consistent assumptions over dynamics improve localization accuracy when compared against a state-of-the-art baseline solution using real-world data from the Oxford Radar RobotCar data set.

cs.RO↗

arXiv2024

ReMoBot: Retrieval-Based Few-Shot Imitation Learning for Mobile Manipulation with Vision Foundation Models

Imitation learning (IL) algorithms typically distill demonstrations into parametric policies to mimic expert behavior. However, with limited data and partial observability, such as in egocentric mobile manipulation, existing methods often struggle to generate accurate actions. To address these challenges, we propose ReMoBot, a few-shot, trajectory-conditioned imitation learning framework that directly Retrieves information from demonstrations to solve Mobile manipulation tasks with ego-centric visual observations. Leveraging vision foundation models, ReMoBot identifies relevant expert demonstrations by combining state-level similarity, history-aware trajectory alignment, and action-sequence consistency to disambiguate perceptually similar observations. The agent then selects appropriate control commands based on these retrieved demonstrations in a fully training-free manner. We evaluate ReMoBot on three mobile manipulation tasks using a Boston Dynamics Spot robot in both simulation and real-world settings. After benchmarking five approaches in simulation, we compare our method with two baselines trained directly on real-world data without sim-to-real transfer. With only 20 demonstrations per task, ReMoBot outperforms the baselines, achieving high success rates in Table Uncover (70%) and Gap Cover (80%), while also showing promising performance on the more challenging Curtain Open task in the real-world setting. Furthermore, ReMoBot generalizes across varying robot positions, object sizes, and material properties, highlighting its robustness in real-world deformable mobile manipulation. Additional details are available at: https://sites.google.com/view/remobot/home

cs.RO↗

arXiv2024

Bayesian Floor Field: Transferring people flow predictions across environments

Mapping people dynamics is a crucial skill for robots, because it enables them to coexist in human-inhabited environments. However, learning a model of people dynamics is a time consuming process which requires observation of large amount of people moving in an environment. Moreover, approaches for mapping dynamics are unable to transfer the learned models across environments: each model is only able to describe the dynamics of the environment it has been built in. However, the impact of architectural geometry on people's movement can be used to anticipate their patterns of dynamics, and recent work has looked into learning maps of dynamics from occupancy. So far however, approaches based on trajectories and those based on geometry have not been combined. In this work we propose a novel Bayesian approach to learn people dynamics able to combine knowledge about the environment geometry with observations from human trajectories. An occupancy-based deep prior is used to build an initial transition model without requiring any observations of pedestrian; the model is then updated when observations become available using Bayesian inference. We demonstrate the ability of our model to increase data efficiency and to generalize across real large-scale environments, which is unprecedented for maps of dynamics.

cs.RO↗

arXiv2024

Do Visual-Language Grid Maps Capture Latent Semantics?

Visual-language models (VLMs) have recently been introduced in robotic mapping using the latent representations, i.e., embeddings, of the VLMs to represent semantics in the map. They allow moving from a limited set of human-created labels toward open-vocabulary scene understanding, which is very useful for robots when operating in complex real-world environments and interacting with humans. While there is anecdotal evidence that maps built this way support downstream tasks, such as navigation, rigorous analysis of the quality of the maps using these embeddings is missing. In this paper, we propose a way to analyze the quality of maps created using VLMs. We investigate two critical properties of map quality: queryability and distinctness. The evaluation of queryability addresses the ability to retrieve information from the embeddings. We investigate intra-map distinctness to study the ability of the embeddings to represent abstract semantic classes and inter-map distinctness to evaluate the generalization properties of the representation. We propose metrics to evaluate these properties and evaluate two state-of-the-art mapping methods, VLMaps and OpenScene, using two encoders, LSeg and OpenSeg, using real-world data from the Matterport3D data set. Our findings show that while 3D features improve queryability, they are not scale invariant, whereas image-based embeddings generalize to multiple map resolutions. This allows the image-based methods to maintain smaller map sizes, which can be crucial for using these methods in real-world deployments. Furthermore, we show that the choice of the encoder has an effect on the results. The results imply that properly thresholding open-vocabulary queries is an open problem.

cs.RO↗

arXiv2023

Object-Oriented Grid Mapping in Dynamic Environments

Grid maps, especially occupancy grid maps, are ubiquitous in many mobile robot applications. To simplify the process of learning the map, grid maps subdivide the world into a grid of cells whose occupancies are independently estimated using measurements in the perceptual field of the particular cell. However, the world consists of objects that span multiple cells, which means that measurements falling onto a cell provide evidence of the occupancy of other cells belonging to the same object. Current models do not capture this correlation and, therefore, do not use object-level information for estimating the state of the environment. In this work, we present a way to generalize the update of grid maps, relaxing the assumption of independence. We propose modeling the relationship between the measurements and the occupancy of each cell as a set of latent variables and jointly estimate those variables and the posterior of the map. We propose a method to estimate the latent variables by clustering based on semantic labels and an extension to the Normal Distributions Transform Occupancy Map (NDT-OM) to facilitate the proposed map update method. We perform comprehensive map creation and localization experiments with real-world data sets and show that the proposed method creates better maps in highly dynamic environments compared to state-of-the-art methods. Finally, we demonstrate the ability of the proposed method to remove occluded objects from the map in a lifelong map update scenario.

cs.RO↗

arXiv2023

Constrained Generative Sampling of 6-DoF Grasps

Most state-of-the-art data-driven grasp sampling methods propose stable and collision-free grasps uniformly on the target object. For bin-picking, executing any of those reachable grasps is sufficient. However, for completing specific tasks, such as squeezing out liquid from a bottle, we want the grasp to be on a specific part of the object's body while avoiding other locations, such as the cap. This work presents a generative grasp sampling network, VCGS, capable of constrained 6 Degrees of Freedom (DoF) grasp sampling. In addition, we also curate a new dataset designed to train and evaluate methods for constrained grasping. The new dataset, called CONG, consists of over 14 million training samples of synthetically rendered point clouds and grasps at random target areas on 2889 objects. VCGS is benchmarked against GraspNet, a state-of-the-art unconstrained grasp sampler, in simulation and on a real robot. The results demonstrate that VCGS achieves a 10-15% higher grasp success rate than the baseline while being 2-3 times as sample efficient. Supplementary material is available on our project website.

cs.RO↗

arXiv2022

LSVL: Large-scale season-invariant visual localization for UAVs

Localization of autonomous unmanned aerial vehicles (UAVs) relies heavily on Global Navigation Satellite Systems (GNSS), which are susceptible to interference. Especially in security applications, robust localization algorithms independent of GNSS are needed to provide dependable operations of autonomous UAVs also in interfered conditions. Typical non-GNSS visual localization approaches rely on known starting pose, work only on a small-sized map, or require known flight paths before a mission starts. We consider the problem of localization with no information on initial pose or planned flight path. We propose a solution for global visual localization on a map at scale up to 100 km2, based on matching orthoprojected UAV images to satellite imagery using learned season-invariant descriptors. We show that the method is able to determine heading, latitude and longitude of the UAV at 12.6-18.7 m lateral translation error in as few as 23.2-44.4 updates from an uninformed initialization, also in situations of significant seasonal appearance difference (winter-summer) between the UAV image and the map. We evaluate the characteristics of multiple neural network architectures for generating the descriptors, and likelihood estimation methods that are able to provide fast convergence and low localization error. We also evaluate the operation of the algorithm using real UAV data and evaluate running time on a real-time embedded platform. We believe this is the first work that is able to recover the pose of an UAV at this scale and rate of convergence, while allowing significant seasonal difference between camera observations and map.

cs.RO↗

arXiv2022

Augmented Environment Representations with Complete Object Models

While 2D occupancy maps commonly used in mobile robotics enable safe navigation in indoor environments, in order for robots to understand and interact with their environment and its inhabitants representing 3D geometry and semantic environment information is required. Semantic information is crucial in effective interpretation of the meanings humans attribute to different parts of a space, while 3D geometry is important for safety and high-level understanding. We propose a pipeline that can generate a multi-layer representation of indoor environments for robotic applications. The proposed representation includes 3D metric-semantic layers, a 2D occupancy layer, and an object instance layer where known objects are replaced with an approximate model obtained through a novel model-matching approach. The metric-semantic layer and the object instance layer are combined to form an augmented representation of the environment. Experiments show that the proposed shape matching method outperforms a state-of-the-art deep learning method when tasked to complete unseen parts of objects in the scene. The pipeline performance translates well from simulation to real world as shown by F1-score analysis, with semantic segmentation accuracy using Mask R-CNN acting as the major bottleneck. Finally, we also demonstrate on a real robotic platform how the multi-layer map can be used to improve navigation safety.

cs.RO↗

arXiv2022

Season-invariant GNSS-denied visual localization for UAVs

Localization without Global Navigation Satellite Systems (GNSS) is a critical functionality in autonomous operations of unmanned aerial vehicles (UAVs). Vision-based localization on a known map can be an effective solution, but it is burdened by two main problems: places have different appearance depending on weather and season, and the perspective discrepancy between the UAV camera image and the map make matching hard. In this work, we propose a localization solution relying on matching of UAV camera images to georeferenced orthophotos with a trained convolutional neural network model that is invariant to significant seasonal appearance difference (winter-summer) between the camera image and map. We compare the convergence speed and localization accuracy of our solution to six reference methods. The results show major improvements with respect to reference methods, especially under high seasonal variation. We finally demonstrate the ability of the method to successfully localize a real UAV, showing that the proposed method is robust to perspective changes.

cs.RO↗