Synthetica – Computer Vision Data Generator
Philips Synthetica is a custom computer vision synthetic training data generator. The tool is build with the Unity 3D Engine and written in C#. Philips Synthetica supports multiple annotation formats among which the COCO format, singleshot/linemod format and keras-yolov3 format. It supports domain-randomization of (grouped) textures, domainrandomization of lights, domain-randomization of backgrounds, mask-image generation and postprocessing noise filters.
Description
Philips Synthetica is a custom computer vision synthetic training data generator. The tool is build with the Unity 3D Engine and written in C#. Philips Synthetica supports multiple annotation formats among which the COCO format, singleshot/linemod format and keras-yolov3 format. It supports domain-randomization of (grouped) textures, domainrandomization of lights, domain-randomization of backgrounds, mask-image generation and postprocessing noise filters.