297 datasets found

D4RL benchmark dataset, which consists of four offline logging datasets, collected by different one or mixed behavior policies.

The Pendulum swing-up task, the agent tries to keep the pendulum upright and balanced under the constraint of keeping away from unsafe angles. In the Reacher task, the robotic...

GraspNet-1Billion is a large-scale real-world grasping dataset containing 190 cluttered grasping scenes and 97,280 RGB-D images captured by 2 kinds of RGB-D cameras from 256...

Grasp-Anything-6D is a large-scale dataset for language-driven 6-DoF grasp detection in 3D point clouds. It consists of 1M point cloud scenes with comprehensive object grasping...

The dataset used in the ACE algorithm for continuous control problems.

The dataset used in the paper for real-time adaptive safety-critical control with Gaussian processes in high-order uncertain models.

D4RL datasets for maze2d-umaze, maze2d-medium, maze2d-large, antmaze-umaze, antmaze-medium, antmaze-large, parking, soccer-sim, and soccer-physical tasks

The FurnitureBench dataset used for the Residual RL for Precise Visual Assembly task.

The dataset used for the Residual RL for Precise Visual Assembly task.

The dataset used in the paper is the Gymnasium MuJoCo benchmark, which is a collection of robotic manipulation tasks.

The dataset used in the paper is the MuJoCo benchmark, which is a collection of robotic manipulation tasks.

The dataset used in this paper consists of command speech, conversation speech, emotion and speaking style-specific speech, Korean speakers' foreign language speech, and Korean...

The dataset is used to train the ClearanceNet model. It contains 106 training and 104 evaluation samples for each environment.

ClearanceNet is a neural network that estimates minimum separation distance, conditioned on joint robot-workspace pose. ClearanceNet is trained using large-scale hyperparameter...

The dataset used in the paper is a Servo Reacher environment, which is a continuous-time control setting.

Generating coherent and natural movement is the key challenge in video generation. This research proposes to condense video generation into a problem of motion generation, to...

The dataset used in the paper is a set of point clouds generated by ORB-SLAM3, a state-of-the-art vision feature-based SLAM system.

The dataset used in the paper is a collection of simulated environments, including blockworld, canada, cart, cartpole, cartpoleBalance, cartpoleParallelDouble,...

Mobile construction tasks in 1/2/3D grid worlds

The Ant, Walker2d, and Hopper datasets are used for testing the proposed method.