Home Categories Code plugins Image Segmentation for Machine Learning

Image Segmentation for Machine Learning

This plugin is designed to streamline the process of generating high-quality datasets for training image segmentation models. With this plugin, you can quickly and easily label meshes in your scene, while the plugin automatically captures the necessary screen captures and creates accompanying label files.

Technical Details

Features:

Labelling of meshes in the scene (works with all kind of meshes including skeletal meshes and spline meshes)
Take simultaneously a screen capture of the scene and save a ground truth file (label).
Setup the number of captures, and the delay between each capture and let it do its work!
Included Python notebook to open your label files and train your own neural network.

Code Modules:

ImgSeg, Developper
ImgSegBP, Developper

Number of Blueprints: 3

Number of C++ Classes: 8

Python notebook: 1

Network Replicated: No

Supported Development Platforms:

Windows: Yes

Mac: No (not tested)

Linux: No (not tested)

Supported Target Build Platforms: Windows

Documentation: https://github.com/Plasma-Lab/Unreal-Image-Segmentation-Doc/blob/main/README.md

Important/Additional Notes: This plugin doesn't work yet with landscape meshes.

Plasma Labs

All Assets by Author

29.60 €

Select

Platforms	Windows 64-bit
UE Versions	5.1 - 5.2
Tags	SYNTHETIC DATA, DEEP LEARNING, DATASET, COMPUTER VISION, IMAGE SEGMENTATION, SEGMENTATION, IMAGE ANNOTATION, SEMANTIC, NEURAL NETWORK, MACHINE LEARNING
Release date	07.05.2023