reinforcement learning for contact rich manipulati

The paper directly addresses reinforcement learning applied to contact-rich manipulation through a residual-RL framework for multi-point object pushing.

The paper explicitly focuses on reinforcement learning for contact-rich manipulation tasks, proposing a framework to ensure safety and robustness during physical interactions.

The paper directly addresses reinforcement learning for whole-body manipulation involving multi-contact interactions with bulky objects, which is a core example of contact-rich manipulation.

The paper directly addresses reinforcement learning for contact-rich dexterous manipulation by proposing an adaptive controller learning framework to bridge the sim-to-real gap.

The paper explicitly focuses on using reinforcement learning for contact-rich, whole-arm manipulation of deformable object clusters.

The paper explicitly focuses on reinforcement learning for robotic pushing, which is a quintessential contact-rich manipulation task, and proposes a method to integrate generated tactile feedback into the RL policy.

The paper directly addresses reinforcement learning for dexterous manipulation in highly contact-rich tasks like screwdriving and nut-bolt fastening.

The paper directly addresses reinforcement learning for a complex contact-rich manipulation task involving cooperative transport via physical contact forces.

The paper is a survey on dexterous robotic manipulation that explicitly identifies reinforcement learning as a key skill-learning framework for these inherently contact-rich tasks.

The paper directly addresses RL for loco-manipulation tasks involving complex physical interactions like pushing and balancing, which are inherently contact-rich.

The paper explicitly focuses on using reinforcement learning for contact-rich loco-manipulation tasks, such as wiping a table, by combining it with trajectory optimization.

The paper directly addresses contact-rich manipulation and compliance using modern policy learning techniques, though it focuses more on imitation learning and flow matching than traditional reinforcement learning.

The paper directly addresses contact-rich manipulation by integrating force signals into VLA models, although it focuses on fine-tuning/imitation learning rather than explicit reinforcement learning.

The paper directly addresses contact-rich manipulation and uses learning-based policies (Diffusion Policy/Behavior Cloning), which is closely related to reinforcement learning in the context of robotic control.

The paper directly addresses contact-rich manipulation using a visual-tactile policy, though it utilizes imitation learning rather than reinforcement learning.

The paper directly addresses contact-rich manipulation and force control using modern policy learning (Diffusion), although it employs imitation learning rather than reinforcement learning.

The paper addresses contact-rich multi-finger manipulation and evaluates its method against a reinforcement learning baseline, although the primary proposed approach uses diffusion models and trajectory optimization.

While the paper focuses on hardware and sensing rather than RL algorithms, it provides a low-cost method for obtaining the contact feedback essential for training and deploying RL policies in contact-rich manipulation.

The paper focuses heavily on contact-rich manipulation and uses learning (BiLSTM), but it employs a control-theoretic approach with policy blending rather than reinforcement learning.

While the paper focuses heavily on contact-rich manipulation and uses learning for estimation, it employs model-based optimization (CQP) rather than reinforcement learning.

The paper focuses on contact-rich manipulation (dexterous grasping), but it employs Model Predictive Control and analytical modeling rather than reinforcement learning.

While the paper focuses on contact-rich manipulation, it proposes a model-based planning and search tree approach rather than reinforcement learning.

While the paper focuses on contact-rich manipulation, it utilizes Vision-Language Models and A* planning rather than reinforcement learning.

While the paper focuses on contact-rich manipulation, it proposes a Model Predictive Control (MPC) and sampling approach rather than Reinforcement Learning.

While the paper focuses on contact-rich manipulation through tactile sensing, it employs an information-theoretic exploration framework rather than reinforcement learning.