Master thesis repository

This repository contains datasets related to real and simulated grasps performed by the NICO robot. The datasets include:

• Datasets of Real and Simulated Robot NICO Grasps
• F5 and STSp grasps implemented by the NeuroNet library.
• UBAL implementation with input data.
• IK Solver program extended to compute coordinates from joint angles.
• Program for rotating 3D points.
• Data converters.
• IR sensor project: A program that enables a robot to perceive the surface of a table using an IR sensor.
• SOM and MSOM implemented in Python.

Data

The data directory contains various data-related files:

• dataset.json: A JSON file containing all grasp data in a structured format. This file is created from text files in the power, precision, and side directories.
• convert_text_file_to_json.py: Python script for converting text files to JSON format.
• histograms: Python scripts for generating histograms.
• power: Text files containing power grasp data.
• precision: Text files containing precision grasp data.
• side: Text files containing side grasp data.

Format of dataset.json

The dataset.json file follows a specific format:

{
    "power": {
        "sequence 0": {
            "position 0": [],
            ...
            "position 15": []
        },
        ...
        "sequence 10": { ... }
    },
    "side": { ... },
    "precision": { ... }
}