arXiv subjects

Davide Scaramuzza

Publications and source records attributed to Davide Scaramuzza.

At least 37 records · Page 2Linked to original sources

arXiv2026

Accelerating Reinforcement Learning via MPC Solver-Gradient Guidance for Weights-varying MPC

In Model Predictive Control (MPC), cost-function weights shape closed-loop behavior, yet changing conditions often make fixed parametrizations suboptimal and motivate context-dependent online adaptation. Learning such policies is difficult because behavior depends implicitly on numerical MPC solutions, producing nonlinear, potentially nonsmooth, long-horizon dependencies on policy parameters. This creates a bias-variance tradeoff: Reinforcement Learning (RL) optimizes realized closed-loop return from environment samples but is sample-inefficient, whereas Gradient-Based Policy Learning (GB-PL) uses low-variance solver gradients from differentiable MPC to optimize surrogate losses on predicted trajectories but can be biased under model mismatch. We propose Solver-Gradient Guided Reinforcement Learning (SG-RL), a solver-sensitivity augmentation for RL-based online MPC cost-weight adaptation. SG-RL keeps sampled closed-loop return as the objective and uses bounded solver-derived gradients as auxiliary guidance to improve stability and sample efficiency. We instantiate SG-RL in Proximal Policy Optimization (PPO) with four modular algorithms that inject solver-gradient guidance into actor-update scaling, policy loss, advantage estimation, and value-function learning. On two full-scale autonomous racing platforms with intentional model mismatch, SG-RL reaches PPO's best closed-loop return with up to 70.6% fewer samples, outperforms GB-PL baselines by at least 54% in closed-loop return, and generalizes zero-shot to unseen environments.

cs.RO↗

arXiv2026

Vision-Based Agile Landing on Turbulent Waters

Autonomous landing of Unmanned Aerial Vehicles on maritime vessels is challenging due to the coupled motion of the vehicle and landing platform in open-sea conditions. This paper presents a reinforcement-learning-based approach for autonomous multirotor landing on moving maritime platforms without requiring explicit platform-state observations or estimation during deployment. The proposed method uses multirotor state measurements together with local visual features, consisting of keypoints and associated descriptors extracted from the landing surface, to predict attitude and thrust commands. These commands are tracked by a conventional low-level controller. The policy is trained in simulation using synthetic keypoints with randomly generated normalized descriptors, enabling zero-shot deployment with different local feature extractors onboard the UAV. We evaluate the method in a realistic simulator and show that it outperforms a state-of-the-art Model Predictive Control baseline under platform motions corresponding to ''Very Rough'' sea conditions. Finally, we perform extensive real-world experiments, demonstrating autonomous onboard landing using two different local feature extractors. To the best of our knowledge, this is the first approach for agile multirotor landing on maritime platforms in turbulent waters that does not rely on an explicit platform-state during deployment.

cs.RO↗

arXiv2026

Low-Latency Event-Based Object Detection with Spatially-Sparse Linear Attention

Event cameras provide sequential visual data with spatial sparsity and high temporal resolution, making them attractive for low-latency object detection. Existing asynchronous event-based neural networks exploit this low-latency advantage by updating predictions event by event, but still suffer from two bottlenecks: recurrent architectures are difficult to train efficiently on long sequences, and improving accuracy often increases per-event computation and latency. Linear attention is appealing because it enables parallel training and recurrent inference. However, its dense state updates make per-event computation scale with the state size, yielding a poor accuracy-efficiency trade-off for object detection, where accurate localization requires fine-grained spatial states. The key challenge is therefore to introduce sparse state activation that exploits the spatial sparsity of events while preserving efficient parallel training. We propose Spatially-Sparse Linear Attention (SSLA), which introduces a mixture-of-spaces state decomposition and a scatter-compute-gather training procedure, enabling state-level sparsity as well as training parallelism. Building on SSLA, we develop an end-to-end asynchronous linear attention model, SSLA-Det, for low-latency event-based object detection. On Gen1 and N-Caltech101, SSLA-Det achieves state-of-the-art accuracy among asynchronous methods, reaching 0.375 mAP and 0.515 mAP, respectively, while reducing per-event computation by over 20 times compared with the strongest prior asynchronous baseline, demonstrating the potential of linear attention for low-latency event-based vision.

cs.CV↗

arXiv2026

ETHEREAL: A 25.6-$μ$s/inf. Low-latency Event-driven Graph-neural-network Processor for High-resolution Vision at the Edge

Dynamic vision sensors (DVS) are enticing candidates to reach the low-latency, sub-ms target of edge-vision applications, as they generate events with a $μ$s-level time resolution. However, using DVS front ends also calls for novel algorithm/hardware back ends capable of efficiently handling streams of sparse spatiotemporal events. While event-driven graph neural networks (EV-GNNs) have emerged as a solution on the algorithmic side that is both accurate and efficient, there is no dedicated hardware to date capable of efficiently supporting their mixed requirements of dense-regular compute operations and sparse-irregular memory accesses. We therefore introduce ETHEREAL, the first EV-GNN processor chip, capable of bridging this gap by means of a neighbor-parallel spline-convolution engine combined with a split-2D/3D memory hierarchy that introduces a novel spatiotemporal event-caching mechanism. Measurement results demonstrate a 25.6$μ$s latency and a 1.6$μ$J energy per end-to-end event-wise inference on the state-of-the art DAGr-GNN workload and VGA-resolution (640x480 pixels) DSEC dataset.

cs.AR↗

arXiv2026

Motion-aware Event Suppression for Event Cameras

Event cameras report asynchronously per-pixel brightness changes with microsecond latency, encoding dynamic visual information as a sparse stream of events. However, their extreme temporal resolution floods perception systems with entangled events from ego-motion and independently moving objects (IMOs), which existing solutions fail to efficiently decouple, relying instead on prohibitive dense 3D reconstructions or limited hand-tuned filters. In this work, we introduce the first framework for Motion-aware Event Suppression, which learns to filter events triggered by IMOs and ego-motion in real time. Our model jointly segments IMOs in the current event stream while predicting their future motion, enabling anticipatory suppression of dynamic events before they occur. Our lightweight architecture achieves 173 Hz inference on consumer-grade GPUs with less than 1 GB of memory usage, outperforming previous state-of-the-art methods on the challenging EVIMO benchmark by 67\% in segmentation accuracy while operating at a 53\% higher inference rate. Moreover, we demonstrate significant benefits for downstream applications: our method accelerates Vision Transformer inference by 83\% via token pruning and improves event-based visual odometry accuracy, reducing Absolute Trajectory Error (ATE) by 13\%.

cs.CV↗

arXiv2026

What Matters in RL-Based Methods for Object-Goal Navigation? An Empirical Study and A Unified Framework

Object-Goal Navigation (ObjectNav) is a key capability for deploying mobile robots in everyday environments such as homes, schools, and workplaces. In this task, an agent must locate an instance of a target object category in previously unseen environments using only onboard perception, requiring the integration of semantic understanding, spatial reasoning, and long-horizon planning. Reinforcement learning (RL) has become a dominant paradigm for ObjectNav, yet modern systems involve numerous design choices across perception modules, policy architectures, and inference-time strategies. The relative impact of these components, however, remains poorly understood. In this work, we present a large-scale empirical study of modular RL-based ObjectNav systems. We decompose the navigation pipeline into three key components: perception, policy, and test-time enhancement, and conduct extensive controlled experiments to analyze their individual contributions. Our results suggest that improvements in perception quality and test-time strategies often yield larger performance gains than policy improvements alone, highlighting the importance of understanding how different components interact within modular navigation systems. Motivated by these findings, we introduce a unified framework for systematically studying modular ObjectNav systems. Guided by our analysis, we build an enhanced system that achieves state-of-the-art performance on the Gibson benchmark, improving SPL by 6.6% and success rate by 2.7% over prior methods. We also introduce a human expert baseline, achieving 98% success, highlighting the significant gap between current RL agents and human-level navigation. Finally, we provide practical insights and design recommendations for each module to help guide future research. Project page: https://honwang0054.github.io/What-matters-in-RL-ObjNav-web/.

cs.RO↗

arXiv2026

Learning Agile Quadrotor Flight in the Real World

Learning-based controllers have achieved impressive performance in agile quadrotor flight but typically rely on massive training in simulation, necessitating accurate system identification for effective Sim2Real transfer. However, even with precise modeling, fixed policies remain susceptible to out-of-distribution scenarios, ranging from external aerodynamic disturbances to internal hardware degradation. To ensure safety under these evolving uncertainties, such controllers are forced to operate with conservative safety margins, inherently constraining their agility outside of controlled settings. While online adaptation offers a potential remedy, safely exploring physical limits remains a critical bottleneck due to data scarcity and safety risks. To bridge this gap, we propose a self-adaptive framework that eliminates the need for precise system identification or offline Sim2Real transfer. We introduce Adaptive Temporal Scaling (ATS) to actively explore platform physical limits, and employ online residual learning to augment a simple nominal model. {Based on the learned hybrid model, we further propose Real-world Anchored Short-horizon Backpropagation Through Time (RASH-BPTT) to achieve efficient and robust in-flight policy updates. Extensive experiments demonstrate that our quadrotor reliably executes agile maneuvers near actuator saturation limits. The system evolves a conservative base policy with a peak speed of 1.9 m/s to 7.3 m/s within approximately 100 seconds of flight time. These findings underscore that real-world adaptation serves not merely to compensate for modeling errors, but as a practical mechanism for sustained performance improvement in aggressive flight regimes.

cs.RO↗

arXiv2026

Regularity and Stability Properties of Selective SSMs with Discontinuous Gating

Selective State-Space Models (SSMs) such as Mamba have become central to long-sequence modeling. Still, their stability is poorly understood: their state-space coefficients are modulated online by a token-dependent gating signal, making the recurrence neither linear time-invariant nor classically nonlinear. We study continuous-time selective SSMs through passivity, dissipativity, and Input-to-State Stability (ISS), explicitly separating the selection signal $x(\cdot)$ from the driving input $u(\cdot)$. We obtain four results: exponential forgetting under strict dissipativity; a canonical $\mathrm{AUC}_{\mathrm{loc}}$ quadratic storage for the frozen-selection subsystem that accommodates discontinuous gating; a parametric LMI together with universal kernel constraints and "irreversible forgetting" under universal quadratic storage; and sufficient conditions for global ISS uniformly over admissible selection schedules. We then bridge to practice by deriving a sampled block LMI for the Mamba selective-scan core, which is used as a differentiable training-time regularizer. Across seven standard time-series datasets and four prediction horizons, the regularizer reduces sampled Mamba-core LMI violations by roughly $92\%$ in $28/28$ pairs at a clean-MSE cost of less than $0.018\%$. It improves internal Mamba passivity and state-norm diagnostics under injected perturbations. Our results turn classical control-theoretic tools into verifiable structural and training criteria for selective SSMs, while honestly scoping which guarantees transfer to a deep selective-scan architecture.

cs.LG↗

arXiv2026

Bridging Performance and Generalization in Reinforcement Learning for Agile Flight

Autonomous drone racing is a fundamentally challenging regime for autonomous aerial robots, requiring time-optimal control while operating under persistent actuation saturation. While reinforcement learning (RL) has achieved human-level performance in this domain, current methods fail to generalize; policies trained on specific environments often crash immediately in unseen configurations. This failure reflects the intrinsic difficulty of zero-shot generalization in agile flight, arising from high-dimensional task variation and the tight coupling between safety and performance at high speeds. Existing approaches that improve generalization impose a substantial cost on flight speed: control policies must significantly degrade performance to achieve even modest levels of generalization. In this work, we propose a framework for zero-shot generalization in agile flight for RL-based drone racing. By combining task-aware switching based on learning progress with a physically informed procedural track generator, the framework produces a fast and robust generalist policy without test-time adaptation. Our method achieves strong zero-shot performance across a wide range of unseen racetracks in the real world, demonstrating a 7.4x improvement in generalization over the state-of-the-art approaches, while maintaining competitive racing speeds. We validate our method's results in both simulation and real-world settings, including a challenging vision-based, end-to-end control setting that operates without explicit state estimation, where all prior approaches fail to generalize.

cs.RO↗

arXiv2026

Continual Robot Policy Learning via Variational Neural Dynamics

Robots deployed in the real world rarely operate under a single fixed dynamics model: wind changes, payloads vary, batteries drain, contacts shift, and hardware wears. Yet most learning-based controllers are trained once and deployed as if learning were complete. This prevents the robot from using deployment experience to further improve task performance. In this work, we propose a continual learning framework that uses real-world experience to improve robot policies under hidden and recurring dynamics. Our method learns a condition-aware dynamics model from real state-action trajectories by combining an analytical physics prior with a neural residual for unmodeled effects. A recurrent encoder infers the current hidden condition from recent interaction, and this estimate conditions both the residual model and the policy. Policy learning is performed via differentiable simulation using diverse learned dynamics sampled from the latent model. At deployment, these sampled conditions are replaced by conditions inferred online from recent real interaction, allowing the policy to recover recurring dynamics by recognition rather than residual re-fitting. Through extensive simulation studies and real-world experiments, we demonstrate that the framework improves policy performance under diverse unobserved disturbances. On real quadrotor trajectory tracking under changing wind, the policy recovers from recurring disturbances in roughly 1s, about 5x faster than online residual re-fitting. It also reduces large-disturbance hover and tracking errors by 65.7% and 53.3% over the state-of-the-art online adaptation approaches

cs.RO↗

arXiv2026

Learning to Throw: Agile and Accurate Cable-Suspended Payload Delivery with a Quadrotor

Quadrotors offer the agility needed to rapidly transport suspended payloads during time-critical applications, including search-and-rescue and medical delivery. While suspended-payload transport and traversal for these missions are well studied, the highly dynamic targeted release of the payload remains comparatively underexplored. State-of-the-art approaches typically rely on model-based trajectory optimization and tracking; however, these methods often yield sub-optimal performance due to conservative feasibility constraints, tracking errors, and the inherent difficulty of analytically modeling flexible rope dynamics. To overcome these limitations, we propose a hybrid simulation framework that couples a high-fidelity analytical quadrotor model with a physics solver for complex rope and payload interactions. By exchanging forces between the two domains at every step, we obtain a physically accurate simulation of the suspended-payload system. Leveraging this environment, we train a deep reinforcement learning (RL) policy that executes agile, accurate payload throws to designated targets. Deployed zero-shot on hardware, our RL policy pushes the boundary of the agility-accuracy trade-off, outperforming the model-based baseline by reducing the landing error by up to 50% and the throw duration by up to 30%. Ablation studies confirm that the coupled simulation is the key enabler of these gains. We further show that the same pipeline trains a policy driven by visual observations rather than an explicit state estimate, achieving accuracy comparable to that of the state-based policy. To accelerate future research in dynamic aerial manipulation, we open-source the simulator to the community upon acceptance.

cs.RO↗

arXiv2026

Neural Events: Discrete Asynchronous Autoencoders for Event-Based Vision

Event cameras capture dynamic scenes with exceptional temporal fidelity by representing them as a continuous stream of microsecond resolution \textit{events}. Each individual event, however, only carries minimal semantic value, merely signaling a localized brightness change. To derive meaningful signals, downstream algorithms need to quickly integrate cues from a potentially massive torrent of low-information events. Current architectures, however, are easily overwhelmed, struggling to balance capturing fine-grained temporal dynamics and maintaining a manageable data throughput. This paper proposes a framework to re-tokenize event streams into a small set of highly informative \textit{neural events}, each representing a local spatio-temporal context window with a discrete learnable code. Every time this code flips, a neural event is triggered, yielding a highly compressed data stream. We demonstrate that, across object detection and classification, networks trained on neural events are on par or surpass the performance of state-of-the-art approaches while reducing the event rate by a factor of 2.0.

cs.CV↗

arXiv2026

Superhuman Safe and Agile Racing through Multi-Agent Reinforcement Learning

Autonomous systems have achieved superhuman performance in isolation or simulation, yet they remain brittle in shared, dynamic real-world spaces. This failure stems from the dominant single-agent paradigm for physical applications, where other actors are ignored or treated as environmental noise, preventing effective coordination. Here we show that multi-agent reinforcement learning provides the essential safety scaffolding required for real-world interaction. Using high-speed quadrotor racing as a high-stakes testbed, we train agents to navigate complex aerodynamic interactions and strategic maneuvering with a variable number of racers. Through league-based self-play, agents evolve sophisticated anticipatory behaviors, including proactive collision avoidance, overtaking, and handling multi-agent physical interactions, including aerodynamic downwash. Our agents outperform a champion-level human pilot in multi-player races at speeds exceeding 22 m/s, while simultaneously reducing collision rates by 50 % compared to state-of-the-art single-agent baselines. Crucially, training with diverse artificial agents enables zero-shot generalization to safer human interaction. These results suggest that the path to robust robotic co-existence lies not in isolated safety constraints, but in the rigorous demands of multi-agent interaction. Multimedia materials are available at: https://rpg.ifi.uzh.ch/marl

cs.RO↗

arXiv2026

Dream to Fly: Model-Based Reinforcement Learning for Vision-Based Drone Flight

Autonomous drone racing has risen as a challenging robotic benchmark for testing the limits of learning, perception, planning, and control. Expert human pilots are able to fly a drone through a race track by mapping pixels from a single camera directly to control commands. Recent works in autonomous drone racing attempting direct pixel-to-commands control policies have relied on either intermediate representations that simplify the observation space or performed extensive bootstrapping using Imitation Learning (IL). This paper leverages DreamerV3 to train visuomotor policies capable of agile flight through a racetrack using only pixels as observations. In contrast to model-free methods like PPO or SAC, which are sample-inefficient and struggle in this setting, our approach acquires drone racing skills from pixels. Notably, a perception-aware behaviour of actively steering the camera toward texture-rich gate regions emerges without the need of handcrafted reward terms for the viewing direction. Our experiments show in both, simulation and real-world flight using a hardware-in-the-loop setup with rendered image observations, how the proposed approach can be deployed on real quadrotors at speeds of up to 9 m/s. These results advance the state of pixel-based autonomous flight and demonstrate that MBRL offers a promising path for real-world robotics research.

cs.RO↗

arXiv2026

Generative Event Pretraining with Foundation Model Alignment

Event cameras provide robust visual signals under fast motion and challenging illumination conditions thanks to their microsecond latency and high dynamic range. However, their unique sensing characteristics and limited labeled data make it challenging to train event-based visual foundation models (VFMs), which are crucial for learning visual features transferable across tasks. To tackle this problem, we propose GEP (Generative Event Pretraining), a two-stage framework that transfers semantic knowledge learned from internet-scale image datasets to event data while learning event-specific temporal dynamics. First, an event encoder is aligned to a frozen VFM through a joint regression-contrastive objective, grounding event features in image semantics. Second, a transformer backbone is autoregressively pretrained on mixed event-image sequences to capture the temporal structure unique to events. Our approach outperforms state-of-the-art event pretraining methods on a diverse range of downstream tasks, including object recognition, segmentation, and depth estimation. Together, VFM-guided alignment and generative sequence modeling yield a semantically rich, temporally aware event model that generalizes robustly across domains.

cs.CV↗

arXiv2026

Maximizing Asynchronicity in Event-based Neural Networks

Event cameras deliver visual data with high temporal resolution, low latency, and minimal redundancy, yet their asynchronous, sparse sequential nature challenges standard tensor-based machine learning (ML). While the recent asynchronous-to-synchronous (A2S) paradigm aims to bridge this gap by asynchronously encoding events into learned features for ML pipelines, existing A2S approaches often sacrifice expressivity and generalizability compared to dense, synchronous methods. This paper introduces EVA (EVent Asynchronous feature learning), a novel A2S framework to generate highly expressive and generalizable event-by-event features. Inspired by the analogy between events and language, EVA uniquely adapts advances from language modeling in linear attention and self-supervised learning for its construction. In demonstration, EVA outperforms prior A2S methods on recognition tasks (DVS128-Gesture and N-Cars), and represents the first A2S framework to successfully master demanding detection tasks, achieving a 0.477 mAP on the Gen1 dataset. These results underscore EVA's potential for advancing real-time event-based vision applications.

cs.LG↗

arXiv2026

Perception-Aware Time-Optimal Planning for Quadrotor Waypoint Flight

Agile quadrotor flight pushes the limits of control, actuation, and onboard perception. While time-optimal trajectory planning has been extensively studied, existing approaches typically neglect the tight coupling between vehicle dynamics, environmental geometry, and the visual requirements of onboard state estimation. As a result, trajectories that are dynamically feasible may fail in closed-loop execution due to degraded visual quality. This paper introduces a unified time-optimal trajectory optimization framework for vision-based quadrotors that explicitly incorporates perception constraints alongside full nonlinear dynamics, rotor actuation limits, aerodynamic effects, camera field-of-view constraints, and convex geometric gate representations. The proposed formulation solves minimum-time lap trajectories for arbitrary racetracks with diverse gate shapes and orientations, while remaining numerically robust and computationally efficient. We derive an information-theoretic position uncertainty metric to quantify visual state-estimation quality and integrate it into the planner through three perception objectives: position uncertainty minimization, sequential field-of-view constraints, and look-ahead alignment. This enables systematic exploration of the trade-offs between speed and perceptual reliability. To accurately track the resulting perception-aware trajectories, we develop a model predictive contouring tracking controller that separates lateral and progress errors. Experiments demonstrate real-world flight speeds up to 9.8 m/s with 0.07 m average tracking error, and closed-loop success rates improved from 55% to 100% on a challenging Split-S course. The proposed system provides a scalable benchmark for studying the fundamental limits of perception-aware, time-optimal autonomous flight.

cs.RO↗

arXiv2026

Event-Aided Sharp Radiance Field Reconstruction for Fast-Flying Drones

Fast-flying aerial robots promise rapid inspection under limited battery constraints, with direct applications in infrastructure inspection, terrain exploration, and search and rescue. However, high speeds lead to severe motion blur in images and induce significant drift and noise in pose estimates, making dense 3D reconstruction with Neural Radiance Fields (NeRFs) particularly challenging due to their high sensitivity to such degradations. In this work, we present a unified framework that leverages asynchronous event streams alongside motion-blurred frames to reconstruct high-fidelity radiance fields from agile drone flights. By embedding event-image fusion into NeRF optimization and jointly refining event-based visual-inertial odometry priors using both event and frame modalities, our method recovers sharp radiance fields and accurate camera trajectories without ground-truth supervision. We validate our approach on both synthetic data and real-world sequences captured by a fast-flying drone. Despite highly dynamic drone flights, where RGB frames are severely degraded by motion blur and pose priors become unreliable, our method reconstructs high-fidelity radiance fields and preserves fine scene details, delivering a performance gain of over 50% on real-world data compared to state-of-the-art methods.

cs.CV↗