The recent great success of Artificial Neural Networks in solving complex problems by approximating the relation (mapping) between input and output spaces motivated many researchers to apply it to machine perception problems (Computer Vision), too. In parallel to that, the great advancement in chip design and microelectronics and the introduction of General-Purpose Graphics Processor Architectures like GPGPU, facilitated training deep neural networks with millions of parameters to achieve state-of-art performance. Unfortunately, training deep learning models requires a great amount of data together with their corresponding annotations or ground truths. Finding, collecting, and annotating suitable data is cumbersome, time-consuming, error-prone, expensive, and subject to privacy issues, just to name a few. The promising solution seems to be hidden in the exceptional state of the art game engines Like Unity, Unreal Engine, and CryEngine. Leveraging the powerful tools of the Unity Game engine, we build a 3D photorealistic virtual world procedurally and at runtime. The system currently supports various computer vision tasks such as semantic segmentation, instance segmentation, depth estimation, and others. It allows researchers, with no computer graphics background, to generate a large-scale synthetic dataset for training or testing their own computer vision models on the fly!