Search arXiv⌕ Search

SEARCH · Search arXiv

Results for “robot learning”

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,729 records · Page 96Linked to original sources

Inferring Smooth Control: Monte Carlo Posterior Policy Iteration with Gaussian Processes

Monte Carlo methods have become increasingly relevant for control of non-differentiable systems, approximate dynamics models and learning from data. These methods scale to high-dimensional spaces and are effective at the non-convex optimizations often seen in robot learning. We look at sample-based methods from the perspective of inference-based control, specifically posterior policy iteration. From this perspective, we highlight how Gaussian noise priors produce rough control actions that are unsuitable for physical robot deployment. Considering smoother Gaussian process priors, as used in episodic reinforcement learning and motion planning, we demonstrate how smoother model predictive control can be achieved using online sequential inference. This inference is realized through an efficient factorization of the action distribution and a novel means of optimizing the likelihood temperature to improve importance sampling accuracy. We evaluate this approach on several high-dimensional robot control tasks, matching the sample efficiency of prior heuristic methods while also ensuring smoothness. Simulation results can be seen at https://monte-carlo-ppi.github.io/.

cs.LG↗

Imitation from Observation: Learning to Imitate Behaviors from Raw Video via Context Translation

Imitation learning is an effective approach for autonomous systems to acquire control policies when an explicit reward function is unavailable, using supervision provided as demonstrations from an expert, typically a human operator. However, standard imitation learning methods assume that the agent receives examples of observation-action tuples that could be provided, for instance, to a supervised learning algorithm. This stands in contrast to how humans and animals imitate: we observe another person performing some behavior and then figure out which actions will realize that behavior, compensating for changes in viewpoint, surroundings, object positions and types, and other factors. We term this kind of imitation learning "imitation-from-observation," and propose an imitation learning method based on video prediction with context translation and deep reinforcement learning. This lifts the assumption in imitation learning that the demonstration should consist of observations in the same environment configuration, and enables a variety of interesting applications, including learning robotic skills that involve tool use simply by observing videos of human tool use. Our experimental results show the effectiveness of our approach in learning a wide range of real-world robotic tasks modeled after common household chores from videos of a human demonstrator, including sweeping, ladling almonds, pushing objects as well as a number of tasks in simulation.

cs.LG↗

World Models: A Comprehensive Survey of Architectures, Methodologies, Reasoning Paradigms, and Applications

World models, internal simulators that learn the structure and dynamics of an environment, have emerged as a central paradigm in the pursuit of artificial general intelligence, enabling agents to predict, plan, and reason within learned representations. Despite rapid progress across reinforcement learning, robotics, autonomous driving, and video generation, the field lacks a unified framework integrating its diverse architectural choices, training methods, reasoning mechanisms, and application settings. This survey addresses that gap with a multi-axis taxonomy organized along four dimensions: (i) architecture, encompassing representation format, dynamics formulation, input modality, learning paradigm, and downstream application; (ii) methodological family, including state-space and recurrent approaches, transformer-based models, diffusion-based generators, physics-informed networks, and language-augmented multimodal systems; (iii) reasoning strategy, covering imagination-based planning, latent policy learning, counterfactual reasoning, and planning under uncertainty; and (iv) application domain, spanning robotics, autonomous driving, video prediction, multimodal agents, reinforcement learning, scientific modeling, medical imaging, educational measurement, and business and finance. Tracing the field from early cognitive-science foundations to milestone systems such as PlaNet, the Dreamer family, MuZero, Sora, Cosmos, and Genie, we examine how these dimensions interact and highlight the recent convergence of chain-of-thought reasoning with world-model imagination. We review evaluation protocols and benchmarks, identify persistent challenges such as compounding prediction errors, sim-to-real transfer, and fragmented evaluation, and outline future directions toward unified multimodal world models, foundation-scale interactive simulators, and safe deployment in safety-critical domains.

cs.LG↗

Survivable Robotic Control through Guided Bayesian Policy Search with Deep Reinforcement Learning

Many robot manipulation skills can be represented with deterministic characteristics and there exist efficient techniques for learning parameterized motor plans for those skills. However, one of the active research challenge still remains to sustain manipulation capabilities in situation of a mechanical failure. Ideally, like biological creatures, a robotic agent should be able to reconfigure its control policy by adapting to dynamic adversaries. In this paper, we propose a method that allows an agent to survive in a situation of mechanical loss, and adaptively learn manipulation with compromised degrees of freedom -- we call our method Survivable Robotic Learning (SRL). Our key idea is to leverage Bayesian policy gradient by encoding knowledge bias in posterior estimation, which in turn alleviates future policy search explorations, in terms of sample efficiency and when compared to random exploration based policy search methods. SRL represents policy priors as Gaussian process, which allows tractable computation of approximate posterior (when true gradient is intractable), by incorporating guided bias as proxy from prior replays. We evaluate our proposed method against off-the-shelf model free learning algorithm (DDPG), testing on a hexapod robot platform which encounters incremental failure emulation, and our experiments show that our method improves largely in terms of sample requirement and quantitative success ratio in all failure modes. A demonstration video of our experiments can be viewed at: https://sites.google.com/view/survivalrl

cs.RO↗

Meta-learning enhanced adaptive robot control strategy for automated PCB assembly

The assembly of printed circuit boards (PCBs) is one of the standard processes in chip production, directly contributing to the quality and performance of the chips. In the automated PCB assembly process, machine vision and coordinate localization methods are commonly employed to guide the positioning of assembly units. However, occlusion or poor lighting conditions can affect the effectiveness of machine vision-based methods. Additionally, the assembly of odd-form components requires highly specialized fixtures for assembly unit positioning, leading to high costs and low flexibility, especially for multi-variety and small-batch production. Drawing on these considerations, a vision-free, model-agnostic meta-method for compensating robotic position errors is proposed, which maximizes the probability of accurate robotic positioning through interactive feedback, thereby reducing the dependency on visual feedback and mitigating the impact of occlusions or lighting variations. The proposed method endows the robot with the capability to learn and adapt to various position errors, inspired by the human instinct for grasping under uncertainties. Furthermore, it is a self-adaptive method that can accelerate the robotic positioning process as more examples are incorporated and learned. Empirical studies show that the proposed method can handle a variety of odd-form components without relying on specialized fixtures, while achieving similar assembly efficiency to highly dedicated automation equipment. As of the writing of this paper, the proposed meta-method has already been implemented in a robotic-based assembly line for odd-form electronic components. Since PCB assembly involves various electronic components with different sizes, shapes, and functions, subsequent studies can focus on assembly sequence and assembly route optimization to further enhance assembly efficiency.

cs.RO↗

Robotic Handling of Compliant Food Objects by Robust Learning from Demonstration

The robotic handling of compliant and deformable food raw materials, characterized by high biological variation, complex geometrical 3D shapes, and mechanical structures and texture, is currently in huge demand in the ocean space, agricultural, and food industries. Many tasks in these industries are performed manually by human operators who, due to the laborious and tedious nature of their tasks, exhibit high variability in execution, with variable outcomes. The introduction of robotic automation for most complex processing tasks has been challenging due to current robot learning policies. A more consistent learning policy involving skilled operators is desired. In this paper, we address the problem of robot learning when presented with inconsistent demonstrations. To this end, we propose a robust learning policy based on Learning from Demonstration (LfD) for robotic grasping of food compliant objects. The approach uses a merging of RGB-D images and tactile data in order to estimate the necessary pose of the gripper, gripper finger configuration and forces exerted on the object in order to achieve effective robot handling. During LfD training, the gripper pose, finger configurations and tactile values for the fingers, as well as RGB-D images are saved. We present an LfD learning policy that automatically removes inconsistent demonstrations, and estimates the teacher's intended policy. The performance of our approach is validated and demonstrated for fragile and compliant food objects with complex 3D shapes. The proposed approach has a vast range of potential applications in the aforementioned industry sectors.

cs.RO↗

Optimal Control-Based Baseline for Guided Exploration in Policy Gradient Methods

In this paper, a novel optimal control-based baseline function is presented for the policy gradient method in deep reinforcement learning (RL). The baseline is obtained by computing the value function of an optimal control problem, which is formed to be closely associated with the RL task. In contrast to the traditional baseline aimed at variance reduction of policy gradient estimates, our work utilizes the optimal control value function to introduce a novel aspect to the role of baseline -- providing guided exploration during policy learning. This aspect is less discussed in prior works. We validate our baseline on robot learning tasks, showing its effectiveness in guided exploration, particularly in sparse reward environments.

cs.LG↗

Efficient and Feasible Robotic Assembly Sequence Planning via Graph Representation Learning

Automatic Robotic Assembly Sequence Planning (RASP) can significantly improve productivity and resilience in modern manufacturing along with the growing need for greater product customization. One of the main challenges in realizing such automation resides in efficiently finding solutions from a growing number of potential sequences for increasingly complex assemblies. Besides, costly feasibility checks are always required for the robotic system. To address this, we propose a holistic graphical approach including a graph representation called Assembly Graph for product assemblies and a policy architecture, Graph Assembly Processing Network, dubbed GRACE for assembly sequence generation. With GRACE, we are able to extract meaningful information from the graph input and predict assembly sequences in a step-by-step manner. In experiments, we show that our approach can predict feasible assembly sequences across product variants of aluminum profiles based on data collected in simulation of a dual-armed robotic system. We further demonstrate that our method is capable of detecting infeasible assemblies, substantially alleviating the undesirable impacts from false predictions, and hence facilitating real-world deployment soon. Code and training data are available at https://github.com/DLR-RM/GRACE.

cs.RO↗

Fuzzy Ensembles of Reinforcement Learning Policies for Robotic Systems with Varied Parameters

Reinforcement Learning (RL) is an emerging approach to control many dynamical systems for which classical control approaches are not applicable or insufficient. However, the resultant policies may not generalize to variations in the parameters that the system may exhibit. This paper presents a powerful yet simple algorithm in which collaboration is facilitated between RL agents that are trained independently to perform the same task but with different system parameters. The independency among agents allows the exploitation of multi-core processing to perform parallel training. Two examples are provided to demonstrate the effectiveness of the proposed technique. The main demonstration is performed on a quadrotor with slung load tracking problem in a real-time experimental setup. It is shown that integrating the developed algorithm outperforms individual policies by reducing the RMSE tracking error. The robustness of the ensemble is also verified against wind disturbance.

cs.RO↗

CTBC: Contact-Triggered Blind Climbing for Wheeled Bipedal Robots with Instruction Learning and Reinforcement Learning

In recent years, wheeled bipedal robots have garnered significant attention due to their exceptional mobility on flat terrain. However, while stair climbing has been achieved in prior studies, these existing methods often suffer from a severe lack of versatility, making them difficult to adapt to varying hardware specifications or diverse complex terrains. To overcome these limitations, we propose a generalized Contact-Triggered Blind Climbing (CTBC) framework. Upon detecting wheel-obstacle contact, the framework triggers a leg-lifting motion integrated with a strongly-guided feedforward trajectory. This allows the robot to rapidly acquire agile climbing skills, significantly enhancing its capability to traverse unstructured environments. Distinct from previous approaches, CTBC demonstrates superior robustness and adaptability, having been validated across multiple wheeled bipedal platforms with different wheel radii and tire materials. Real-world experiments demonstrate that, relying solely on proprioceptive feedback, the proposed framework enables robots to achieve reliable and continuous climbing over obstacles well beyond their wheel radius.

cs.RO↗

GPUSimBench: Towards Scalable and Reliable GPU-Accelerated Simulators in Embodied AI

Data-driven embodied AI is rapidly transitioning into a paradigm that scales training through massively parallel simulation, where GPU-accelerated simulators serve as the foundational data infrastructure. However, as computational throughput scales, the underlying trade-offs between parallel efficiency, physical fidelity, and execution determinism remain largely unexamined, hindering the development of reliable robot learning. In this paper, we expose the hidden limits of mainstream GPU-based robotic simulators (e.g., Isaac Lab, Genesis) by introducing GPUSimBench, which focuses on scalability, physical consistency, and computational determinism. First, GPUSimBench establishes a physical grounding evaluation with a controlled inclined-plane task, quantifying the distributional alignment between simulated dynamics and their real-world counterparts. Second, we benchmark parallel scalability by measuring throughput and memory footprints across scaling environment counts. Crucially, beyond standard performance metrics, we unveil and quantify the inherent non-determinism introduced by GPU-batched execution, characterized by significant run-to-run and inter-environment variability even under identical initial conditions. Finally, we identify four empirical regimes of stochasticity within current simulator stacks, highlighting that unbounded scaling can compromise reproducibility without explicit constraints.

cs.RO↗

Robot-WATCHDOG : failure detection through object-centric graph representation

Reliable real-time failure detection is critical for deploying learned robotic policies, yet defining failures explicitly is infeasible in open-world settings. This motivates failure detection from nominal demonstrations alone. Existing methods, however, struggle to balance robustness to benign variations with the ability to capture semantic and interaction-level errors. We introduce Robot-WATCHDOG, a self-supervised framework that models manipulation scenes as object-centric spatio-temporal graphs. Building on this representation, we investigate two complementary failure scores: Trajectory Projection and Correlation (TPC), a non-parametric measure of geometric deviation and relational consistency, and GnnT, a Graph Attention Network with a Transformer that predicts future object states. We evaluate both on a public benchmark and on a new dataset of semantically complex manipulation tasks, across six tasks and ten random seeds. Both scores outperform image-based failure detectors, particularly on semantic and interaction-level failures. TPC provides the strongest performance in AUPR and MCC on both dataset averages, with low variance across seeds. Qualitative results are made available at https://cea-list.github.io/Robot-WATCHDOG/.

cs.RO↗

Learning from Demonstration with Failure Awareness for Safe Robot Navigation

Learning from demonstration is widely used for robot navigation, yet it suffers from a fundamental limitation: demonstrations consist predominantly of successful behaviors and provide limited coverage of unsafe states. This limitation leads to poor safety when the robot encounters scenarios beyond the demonstration distribution. Failure experiences, such as collisions, contain essential information about unsafe regions, but remain underutilized. The key difficulty lies in the fact that failure data do not provide valid guidance for action imitation, and their naive incorporation into policy learning often degrades performance. We address this challenge by proposing a failure-aware learning framework that explicitly decouples the roles of success and failure data. In this framework, failure experiences are used to shape value estimation in hazardous regions, while policy learning is restricted to successful demonstrations. This separation enables the effective use of failure data without corrupting policy behavior. We implement this design within an offline reinforcement learning (RL) setting and evaluate it in both simulation and real-world environments. The results show that our framework consistently reduces collision rates while preserving the task success rate, and demonstrate strong generalization across different environments and robot platforms.

cs.RO↗

Uncertainty Aware Learning from Demonstrations in Multiple Contexts using Bayesian Neural Networks

Diversity of environments is a key challenge that causes learned robotic controllers to fail due to the discrepancies between the training and evaluation conditions. Training from demonstrations in various conditions can mitigate---but not completely prevent---such failures. Learned controllers such as neural networks typically do not have a notion of uncertainty that allows to diagnose an offset between training and testing conditions, and potentially intervene. In this work, we propose to use Bayesian Neural Networks, which have such a notion of uncertainty. We show that uncertainty can be leveraged to consistently detect situations in high-dimensional simulated and real robotic domains in which the performance of the learned controller would be sub-par. Also, we show that such an uncertainty based solution allows making an informed decision about when to invoke a fallback strategy. One fallback strategy is to request more data. We empirically show that providing data only when requested results in increased data-efficiency.

cs.RO↗

Grasp Anything: Combining Teacher-Augmented Policy Gradient Learning with Instance Segmentation to Grasp Arbitrary Objects

Interactive grasping from clutter, akin to human dexterity, is one of the longest-standing problems in robot learning. Challenges stem from the intricacies of visual perception, the demand for precise motor skills, and the complex interplay between the two. In this work, we present Teacher-Augmented Policy Gradient (TAPG), a novel two-stage learning framework that synergizes reinforcement learning and policy distillation. After training a teacher policy to master the motor control based on object pose information, TAPG facilitates guided, yet adaptive, learning of a sensorimotor policy, based on object segmentation. We zero-shot transfer from simulation to a real robot by using Segment Anything Model for promptable object segmentation. Our trained policies adeptly grasp a wide variety of objects from cluttered scenarios in simulation and the real world based on human-understandable prompts. Furthermore, we show robust zero-shot transfer to novel objects. Videos of our experiments are available at \url{https://maltemosbach.github.io/grasp_anything}.

cs.RO↗

Multiscale Sensor Fusion and Continuous Control with Neural CDEs

Though robot learning is often formulated in terms of discrete-time Markov decision processes (MDPs), physical robots require near-continuous multiscale feedback control. Machines operate on multiple asynchronous sensing modalities, each with different frequencies, e.g., video frames at 30Hz, proprioceptive state at 100Hz, force-torque data at 500Hz, etc. While the classic approach is to batch observations into fixed-time windows then pass them through feed-forward encoders (e.g., with deep networks), we show that there exists a more elegant approach -- one that treats policy learning as modeling latent state dynamics in continuous-time. Specifically, we present 'InFuser', a unified architecture that trains continuous time-policies with Neural Controlled Differential Equations (CDEs). InFuser evolves a single latent state representation over time by (In)tegrating and (Fus)ing multi-sensory observations (arriving at different frequencies), and inferring actions in continuous-time. This enables policies that can react to multi-frequency multi sensory feedback for truly end-to-end visuomotor control, without discrete-time assumptions. Behavior cloning experiments demonstrate that InFuser learns robust policies for dynamic tasks (e.g., swinging a ball into a cup) notably outperforming several baselines in settings where observations from one sensing modality can arrive at much sparser intervals than others.

cs.RO↗

Learning Real-World Action-Video Dynamics with Heterogeneous Masked Autoregression

We propose Heterogeneous Masked Autoregression (HMA) for modeling action-video dynamics to generate high-quality data and evaluation in scaling robot learning. Building interactive video world models and policies for robotics is difficult due to the challenge of handling diverse settings while maintaining computational efficiency to run in real time. HMA uses heterogeneous pre-training from observations and action sequences across different robotic embodiments, domains, and tasks. HMA uses masked autoregression to generate quantized or soft tokens for video predictions. \ourshort achieves better visual fidelity and controllability than the previous robotic video generation models with 15 times faster speed in the real world. After post-training, this model can be used as a video simulator from low-level action inputs for evaluating policies and generating synthetic data. See this link https://liruiw.github.io/hma for more information.

cs.RO↗

Early or Late Fusion Matters: Efficient RGB-D Fusion in Vision Transformers for 3D Object Recognition

The Vision Transformer (ViT) architecture has established its place in computer vision literature, however, training ViTs for RGB-D object recognition remains an understudied topic, viewed in recent literature only through the lens of multi-task pretraining in multiple vision modalities. Such approaches are often computationally intensive, relying on the scale of multiple pretraining datasets to align RGB with 3D information. In this work, we propose a simple yet strong recipe for transferring pretrained ViTs in RGB-D domains for 3D object recognition, focusing on fusing RGB and depth representations encoded jointly by the ViT. Compared to previous works in multimodal Transformers, the key challenge here is to use the attested flexibility of ViTs to capture cross-modal interactions at the downstream and not the pretraining stage. We explore which depth representation is better in terms of resulting accuracy and compare early and late fusion techniques for aligning the RGB and depth modalities within the ViT architecture. Experimental results in the Washington RGB-D Objects dataset (ROD) demonstrate that in such RGB -> RGB-D scenarios, late fusion techniques work better than most popularly employed early fusion. With our transfer baseline, fusion ViTs score up to 95.4% top-1 accuracy in ROD, achieving new state-of-the-art results in this benchmark. We further show the benefits of using our multimodal fusion baseline over unimodal feature extractors in a synthetic-to-real visual adaptation as well as in an open-ended lifelong learning scenario in the ROD benchmark, where our model outperforms previous works by a margin of >8%. Finally, we integrate our method with a robot framework and demonstrate how it can serve as a perception utility in an interactive robot learning scenario, both in simulation and with a real robot.

cs.CV↗