“Good Robot!”: Efficient Reinforcement Learning for Multi-Step Visual Tasks with Sim to Real Transfer

Hundt, Andrew; Killeen, Benjamin D.; Greene, Nicholas D. E.; Wu, Hongtao; Kwon, Heeyeon; Paxton, Chris; Hager, Gregory D.

doi:10.1109/lra.2020.3015448

Cited by 59 publications

(51 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In practice, three elements are vital to help transfer learning improve its performance than building and training a network [31] from scratch: (i) Successful PTM can help the user remove hyper-parameter tuning; (ii) The initial layers in PTM can be thought of as feature descriptors, which extract low-level features, e.g., tints, edges, blobs, shades, and textures; (iii) The target model may only need to re-train the last several layers of the pre-trained model, since we believe the last several layers carry out the complex identification tasks. The basic idea of transfer learning is shown in Fig.…”

Section: Methodsmentioning

confidence: 99%

COVID-19 classification by CCSHNet with deep fusion using transfer learning and discriminant correlation analysis

et al. 2021

View full text Add to dashboard Cite

Highlights We proposed a novel (L, 2) transfer feature learning (L2TFL) approach. L2TFL can elucidate the optimal layers to be removed prior to selection. We developed a novel selection algorithm of pretrained network for fusion approach. SAPNF can determine the best two pretrained models for fusion. We introduced a deep CCT fusion discriminant correlation analysis fusion method.

show abstract

Section: Methodsmentioning

confidence: 99%

COVID-19 classification by CCSHNet with deep fusion using transfer learning and discriminant correlation analysis

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Gripping is also an important challenge in the robotic field, as performing robust gripping tasks in various situations is studied in many places [55], [56]. Unlike the above scenarios, gripping scenarios induce the combined contact situation between both dynamically moving hard and soft parts, which should be handled in a D-contact manner.…”

Section: Grippingmentioning

confidence: 99%

Large-Dimensional Multibody Dynamics Simulation Using Contact Nodalization and Diagonalization

Lee¹,

Lee²,

Lee³

2022

Preprint

View full text Add to dashboard Cite

We propose a novel multibody dynamics simulation framework that can efficiently deal with large-dimensionality and complementarity multi-contact conditions. Typical contact simulation approaches perform contact impulse-level fixed-point iteration (IL-FPI), which has high time-complexity from large-size matrix inversion and multiplication, as well as susceptibility to illconditioned contact situations. To circumvent this, we propose a novel framework based on velocity-level fixed-point iteration (VL-FPI), which, by utilizing a certain surrogate dynamics and contact nodalization (with virtual nodes), can achieve not only intercontact decoupling but also their inter-axes decoupling (i.e., contact diagonalization). This then enables us to one-shot/parallelsolve the contact problem during each VL-FPI iteration-loop, while the surrogate dynamics structure allows us to circumvent large-size/dense matrix inversion/multiplication, thereby, significantly speeding up the simulation time with improved convergence property. We theoretically show that the solution of our framework is consistent with that of the original problem and, further, elucidate mathematical conditions for the convergence of our proposed solver. Performance and properties of our proposed simulation framework are also demonstrated and experimentallyvalidated for various large-dimensional/multi-contact scenarios including deformable objects.

show abstract

“…One of the approaches is focused on unscrewing operations in robotic disassembly of electronic waste using the Q-Learning method [16], other approaches have used a robotic arm to achieve a goal using the Deep Reinforcement Learning method DQN (Deep Q-Network) [5], TRPO(Trust Region Policy Optimization) [22] and tested the result of the experiment in a real application. Some approaches use 2D/3D cameras and some other sensors to observe the robotic environment [5,10,12], others use only dynamic simulation with the specified environment [13,16] or use real-time robot learning techniques [19].…”

Section: Related Workmentioning

confidence: 99%

Comparison of Multiple Reinforcement Learning and Deep Reinforcement Learning Methods for the Task Aimed at Achieving the Goal

Parak

Matoušek

2021

mendel

View full text Add to dashboard Cite

Reinforcement Learning (RL) and Deep Reinforcement Learning (DRL) methods are a promising approach to solving complex tasks in the real world with physical robots. In this paper, we compare several reinforcement learning (Q-Learning, SARSA) and deep reinforcement learning (Deep Q-Network, Deep Sarsa) methods for a task aimed at achieving a specific goal using robotics arm UR3. The main optimization problem of this experiment is to find the best solution for each RL/DRL scenario and minimize the Euclidean distance accuracy error and smooth the resulting path by the Bézier spline method. The simulation and real word applications are controlled by the Robot Operating System (ROS). The learning environment is implemented using the OpenAI Gym library which uses the RVIZ simulation tool and the Gazebo 3D modeling tool for dynamics and kinematics.

show abstract

“Good Robot!”: Efficient Reinforcement Learning for Multi-Step Visual Tasks with Sim to Real Transfer

Cited by 59 publications

References 23 publications

COVID-19 classification by CCSHNet with deep fusion using transfer learning and discriminant correlation analysis

COVID-19 classification by CCSHNet with deep fusion using transfer learning and discriminant correlation analysis

Large-Dimensional Multibody Dynamics Simulation Using Contact Nodalization and Diagonalization

Comparison of Multiple Reinforcement Learning and Deep Reinforcement Learning Methods for the Task Aimed at Achieving the Goal

Contact Info

Product

Resources

About