Improved EBM training yields better samples and robust performance.
problem Training energy-based models (EBMs) on high-dimensional data.
method Scaled MCMC-based training techniques for continuous neural networks.
result Achieved state-of-the-art performance on various tasks.
Robot learns to imitate human interactions through deep learning.
problem Teaching robots to coordinate actions with human partners.
method Deep learning framework for motion embedding, prediction, and trajectory generation.
result Importance of predictive and adaptive components for successful imitation.
Deep Reinforcement Learning has shown tremendous success in solving several games and tasks in robotics. However, unlike humans, it generally requires a lot of training instances. Trajectories imitating to solve the task at hand can help to increase sample-efficiency of deep RL methods. In this paper, we present a simp…
New method constructs geometric flat outputs for robotic systems using symmetry.
problem Finding flat outputs for arbitrary robotic systems remains an open question.
method Employing symmetry directly to construct a flat output.
result Demonstrated geometric flat outputs for various robotic systems.
Robot learns user preferences from brain signals.
problem Decoding user preferences for robot motions from brain signals.
method Proposes a novel approach using electroencephalography to decode user preferences from brain signals.
result Brain signals can reliably infer user preferences for robot trajectories.
Deep model predicts robot trajectories in real time.
problem Real-time robot trajectory prediction with low latency.
method Deep conditional generative model trained with SGDVB.
result More accurate long-term predictions with lower latency.
New model accounts for continuous human trajectories in robotics.
problem Inaccurate probabilistic models of human behavior in robotics.
method Developed a new probabilistic model that considers distances between continuous trajectories.
result The new model outperforms existing models in explaining human behavior and improving robot inference.
DIGIT is a low-cost tactile sensor for in-hand manipulation.
problem Difficulty in sensing contact forces limits robotic manipulation.
method DIGIT miniaturizes and improves a vision-based tactile sensor.
result DIGIT enables better control of interactions with the environment.
One of the open challenges in designing robots that operate successfully in the unpredictable human environment is how to make them able to predict what actions they can perform on objects, and what their effects will be, i.e., the ability to perceive object affordances. Since modeling all the possible world interactio…
Transfer learning improves dynamic model transfer for soft robotic hands.
problem Improving dynamic model transfer for underactuated soft robotic hands.
method Derive an upper bound on the Lyapunov exponent and use it to improve transfer.
result Significant improvement over traditional fine-tuning methods.
Reinforcement learning optimizes robot trajectories for unknown dynamics.
problem Optimizing robot trajectories for systems with unknown dynamics.
method Curriculum learning with reinforcement learning to generate smooth trajectories.
result Reinforcement learning agent outperforms PID controllers in trajectory tracking.
Robot solves Rubik's cube using simulation-trained models.
problem Solving Rubik's cube with a robot hand.
method Automatic domain randomization (ADR) and a custom robot platform.
result Control policies and vision state estimators trained with ADR exhibit improved sim2real transfer.
Hierarchical RL system for robotic manipulation with explainable decision-making.
problem Interpretability of robot decision-making for human operators.
method Dot-to-Dot: Hierarchical Deep Reinforcement Learning.
result Efficient learning of complex actions/states by low-level agent and interpretable high-level representation.
The paper presents a method to reduce arm motion complexity for prosthetics and robotics.
problem Reducing the complexity of human arm motions for robotic and prosthetic control.
method Data-driven techniques including DTW, DBA, Ward's distance, batch-DTW, and fPCA.
result Representative motion clusters and averages for different arm DOF levels.
In Hindsight Experience Replay (HER), a reinforcement learning agent is trained by treating whatever it has achieved as virtual goals. However, in previous work, the experience was replayed at random, without considering which episode might be the most valuable for learning. In this paper, we develop an energy-based fr…
Paper proposes a method for optimizing local policies for trajectory-centric reinforcement learning.
problem Challenges in global policy optimization for non-linear systems and poor performance of open-loop trajectory optimization.
method Formulates trajectory optimization and local policy synthesis as a single optimization problem and solves it as a nonlinear programming instance.
result Demonstrates improved performance of the proposed technique under simplifying assumptions.
A new Bayesian MBRL method improves performance in robotics tasks.
problem Enhancing model-based reinforcement learning with uncertainty.
method Introduces variational inference MPC and probabilistic action ensembles with trajectory sampling (PaETS).
result Consistently improves performance on challenging locomotion tasks.
Paper teaches robots to play piano with touch and learning.
problem Teaching robots to play piano with touch and emotion.
method Reinforcement learning from scratch with touch-augmented reward and curriculum.
result Robots can play piano with correct key positions and various requirements.
Develops a method to plan exploration that learns strong policies with fewer samples.
problem Lack of efficient exploration in reinforcement learning for real-world tasks.
method Plans an action sequence that maximizes information gain about the optimal trajectory.
result 2x fewer samples than exploration baselines and 200x fewer than model-free methods.
The paper learns robot skills from demonstrations without supervision.
problem Discovering robotic options from unlabelled demonstrations.
method Temporal variational inference for latent variable learning.
result The framework can learn options across multiple datasets.
This work creates a system for understanding human movement in spaces.
problem Simplify communication and interaction between robots and humans in spatial tasks.
method Uses unsupervised learning with neural autoencoding to learn continuous representations of spatio-temporal trajectory data.
result Proposes a method to form prototypical representations of movement based on spatial context.
New algorithm reduces sample complexity for imitation learning.
problem High sample complexity limits deployment of adversarial imitation algorithms.
method Trajectory-centric reinforcement learning ideas.
result Improved learning rate and efficiency in imitation tasks.
This research evaluates learning models for bionic robots, focusing on transfer function identification.
problem Developers need guidance on selecting and constructing transfer functions for bionic robots.
method Comprehensive evaluation strategy including data collection, learning model selection, comparative analysis, and transfer function identification.
result A framework for effectively dealing with multi-input multi-output robotic data.
This work tackles real-world robotic reinforcement learning challenges.
problem Limited success of reinforcement learning in real-world robotics.
method Proposes a system for autonomous real-world learning without instrumentation.
result Demonstrates a complete system that learns without human intervention.
This work presents a methodology to design trajectory tracking feedback control laws, which embed non-parametric statistical models, such as Gaussian Processes (GPs). The aim is to minimize unmodeled dynamics such as undesired slippages. The proposed approach has the benefit of avoiding complex terramechanics analysis …
New method learns state embeddings from demonstrations for improved reinforcement learning.
problem Difficult relationship between observed state and useful policy actions in dynamic problems.
method Variational framework for learning state embeddings that optimize trajectory linearity.
result Learning embedding spaces improves policy gradient reinforcement learning performance.
Deep learning is an established framework for learning hierarchical data representations. While compute power is in abundance, one of the main challenges in applying this framework to robotic grasping has been obtaining the amount of data needed to learn these representations, and structuring the data to the task at ha…
Diffusion models enhance robotic manipulation through probabilistic multi-modal learning.
problem Enhancing robotic manipulation through robust and multi-modal learning.
method Probabilistic diffusion models integrating imitation and reinforcement learning.
result Diffusion models improve grasp learning, trajectory planning, and data augmentation in robotics.
In Reinforcement Learning (RL), an agent explores the environment and collects trajectories into the memory buffer for later learning. However, the collected trajectories can easily be imbalanced with respect to the achieved goal states. The problem of learning from imbalanced data is a well-known problem in supervised…
Using movement primitive libraries is an effective means to enable robots to solve more complex tasks. In order to build these movement libraries, current algorithms require a prior segmentation of the demonstration trajectories. A promising approach is to model the trajectory as being generated by a set of Switching L…
Most policy search algorithms require thousands of training episodes to find an effective policy, which is often infeasible with a physical robot. This survey article focuses on the extreme other end of the spectrum: how can a robot adapt with only a handful of trials (a dozen) and a few minutes? By analogy with the wo…
ROBEL platform accelerates reinforcement learning with low-cost robots.
problem Accelerating reinforcement learning research in robotics.
method Open-source platform of cost-effective robots for real-world reinforcement learning.
result Robots D'Claw and D'Kitty facilitate learning dexterous manipulation and agile locomotion tasks.
Paper learns domain randomization distributions for robust robot policies.
problem Finding good domain randomization parameters for simulation without real data.
method Gradient-based search methods to learn domain randomization distribution.
result Improvements in jump-start and asymptotic performance when transferring policies.
ORIL learns a reward function from unlabeled data to improve robot learning.
problem Leveraging unlabeled data for robot learning.
method ORIL learns a reward function from demonstrator and unlabeled trajectories, annotates data, and trains an agent via offline reinforcement learning.
result ORIL consistently outperforms BC agents on various robotic tasks.
RL controls small soccer robots in a real league, beating human-designed policies.
problem Training robots to play complex, real-world sports.
method Sim-to-Real RL approach, training in simulated environment, applying to real-world robots.
result Robots learned policies to compete effectively, beating human-designed strategies.
This paper evaluates various representations for robotics tasks, improving performance in lifting, stacking, and pushing.
problem Improving data-efficiency in reinforcement learning for robotics with limited data.
method Systematic evaluation of common representations in three robotics tasks: lifting, stacking, and pushing.
result Some representations can perform as well as simulator states as agent inputs, challenging common intuitions.
Active IRL selects optimal human demonstrations for learning AI preferences.
problem Costly human demonstrations in IRL for autonomous systems.
method Information-theoretic acquisition function for selecting informative human demonstrations.
result Efficiently reduces human effort in learning AI preferences.
Paper introduces OTR for efficient offline RL in surgical robotics.
problem Lack of annotated datasets for offline RL in surgical robotics.
method OTR algorithm using Optimal Transport to assign rewards to unlabeled trajectories.
result OTR enables efficient policy learning from large datasets without handcrafted rewards.
Paper provides closed-form time derivatives for rigid body systems.
problem Need for time derivatives of equations of motion in robotics.
method Lie group formulation for rigid body systems to derive closed-form derivatives up to second-order.
result Closed-form equations provide direct insight into system dynamics.
Measuring similarities between unlabeled time series trajectories is an important problem in domains as diverse as medicine, astronomy, finance, and computer vision. It is often unclear what is the appropriate metric to use because of the complex nature of noise in the trajectories (e.g. different sampling rates or out…
Symmetry augmentation speeds up learning in robotics tasks.
problem Learning efficiency in robotics tasks with limited data.
method Data augmentation using symmetry in the quadruped domain of DeepMind control suite.
result Agent learns faster with augmented symmetry experiences.
Proposes deep optimal feedback control for continuous-time systems with action constraints.
problem Learning optimal feedback control laws for robotic applications.
method Exploits Hamilton-Jacobi-Bellman equation and deep differential networks to learn optimal value function and feedback policy.
result Enables learning an optimal feedback control law that generates an optimal trajectory from any point in state-space without replanning.
SAVED safely learns robot tasks with sparse rewards.
problem Challenges in reinforcement learning for robotics, especially sparse rewards and complex constraints.
method SAVED uses supervision to constrain exploration and learn efficiently, handling complex constraints.
result SAVED outperforms state-of-the-art methods in success rate, constraint satisfaction, and sample efficiency.
Swept Volume (SV), the volume displaced by an object when it is moving along a trajectory, is considered a useful metric for motion planning. First, SV has been used to identify collisions along a trajectory, because it directly measures the amount of space required for an object to move. Second, in sampling-based moti…
This thesis tackles learning reward functions from human comparative feedback.
problem Designing reward functions for complex tasks is challenging and humans often provide suboptimal demonstrations.
method Proposes learning reward functions from comparative feedback (pairwise comparisons, best-of-many choices, rankings, scaled comparisons) and active learning techniques.
result Demonstrates the effectiveness of learning reward functions from comparative feedback in various domains.
A new method for robot manipulation tasks using imagined object goals.
problem Learning robot manipulation in sparse reward environments.
method Train objects to reach target positions, then use predictions to create a curriculum of tasks.
result Higher success rates in challenging learning scenarios compared to alternatives.
SDM Policy accelerates inference for robotic tasks while maintaining high action quality.
problem Prolonged inference times in diffusion-based policies for high-frequency control tasks.
method Two-stage optimization: score matching and distribution matching; dual-teacher mechanism.
result 6x inference speedup with state-of-the-art action quality.
This research enhances exploration in DDPG using latent trajectory optimization.
problem Limited exploration in DDPG with deterministic policies.
method Model-based trajectory optimization for exploration in DDPG, using a learned deep dynamics model.
result Improved performance in continuous control tasks, especially with sparse rewards and images.