Learning to grasp under uncertainty

By Freek Stulp, Evangelos A. Theodorou, Jonas Buchli, Stefan Schaal. 2011

98 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Control and Systems Engineering, Reinforcement Learning, Robotics & Control

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Sergey Levine
2. Pieter Abbeel
3. Chelsea Finn
4. Shimon Whiteson
5. J. Urgen Schmidhuber
6. Martin A. Riedmiller
7. Mrinal Kalakrishnan
8. Thomas Lampe

In Cognitive Neuroscience

1. Sergey Levine
2. Martin A. Riedmiller
3. Thomas Lampe
4. Peter Pastor
5. Freek Stulp
6. Olivier Sigaud
7. Mrinal Kalakrishnan
8. Shimon Whiteson

In Control and Systems Engineering

1. Sergey Levine
2. Pieter Abbeel
3. Chelsea Finn
4. Mrinal Kalakrishnan
5. Peter Pastor
6. Jan Peters
7. Thomas Lampe
8. Martin A. Riedmiller

In Reinforcement Learning

1. Sergey Levine
2. Pieter Abbeel
3. Chelsea Finn
4. Shimon Whiteson
5. Jan Peters
6. Mrinal Kalakrishnan
7. Yevgen Chebotar
8. Martin A. Riedmiller

In Robotics & Control

1. Sergey Levine
2. Mrinal Kalakrishnan
3. Pieter Abbeel
4. Thomas Lampe
5. Peter Pastor
6. Martin A. Riedmiller
7. Chelsea Finn
8. Freek Stulp

Citations in the corpus

• A General Framework for Uncertainty Estimation in Deep Learning (2019, background): “Also for the task of robot manipulation, uncertainty estimation was shown to play a fundamental role to increase the learning efficiency and guarantee the manipulator safety [21], [22].”
• Path integral guided policy search (2016, methodology): “PI(2) has been successfully used to learn parameters of trajectory-centric policies such as dynamic movement primitives [19] in grasping [20], pick-and-place [21] and variable impedance control tasks [22].”
• Path Integral Policy Improvement with Covariance Matrix Adaptation (2012, background): “Given the ability of PI 2 to learn complex, high-dimensional tasks on real robots (Stulp et al., 2011), we are con-fident that PI 2 -CMA can also successfully be applied to such tasks.”
• Transient Non-stationarity and Generalisation in Deep Reinforcement Learning (2020, background): “, 2017), inject noise into the environment (Stulp et al., 2011; Lee et al., 2020), incorporate inductive biases in the architecture (Kansky et al.”
• A survey of uncertainty in deep neural networks (2021, background): “[325], [28], [326], [327] demonstrated uncertainty-aware, real world application of a reinforcement learning algorithm for robotics, while [328], [329] proposed to leverage spatial information, on top of MC-dropout.”
• Acquiring visual servoing reaching and grasping skills using neural reinforcement learning (2013, methodology): “Reaching and grasping tasks have been extensively covered using combinations of policy iteration techniques and imitation learning [1], [7], [21], [22].”
• End-to-end dexterous manipulation with deliberate interactive estimation (2012, background): “Recent work has applied reinforcement learning with probabilistic distributions of object pose to determine grasps that are resistant to state estimation errors [21].”
• Learning motion primitive goals for robust manipulation (2011, methodology): “In contrast to the work in [17], here we • use a simplified version of PI2 • apply simultaneous learning of goal and shape parameters, rather than learning shape alone.”
• A data-driven statistical framework for post-grasp manipulation (2014, background): “[41] learned motion primitives to optimize the chance of grasping an object with Gaussian uncertainty on its location.”