Dataset to Build Neural Network-Based Inverse Kinematics for NAO Robot Arms

The dataset to build the neural network-based inverse kinematics for NAO robot arms consists of input-output data pairs. The input data contains the end-effector position and orientation, while the output data contains a set of joint angular positions. The dataset that we published in this data repository contains 10000 data. For further applications, this dataset was split into the training dataset, validation dataset, and testing dataset. From 10000 data, 60% of data was allocated for the training dataset, 20% of data for the validation dataset, and the remaining 20% of data for the testing dataset. The training dataset is used to the train neural network. The validation dataset is utilized to validate the neural network’s performance during the training process. Meanwhile, the testing dataset is employed after the training process to test the performance of the trained neural network. It should be noted, this dataset can be applied for NAO H25 v3.3 or later.