All the images shown below are generated using NYX-IMAGEN, our latest in-house T2I (Text to Image) pipeline.
This pipeline utilizes what we call "multi-layer reinforcement stable diffusion", a technical approach at inference runtime with the purpose of consistent style image generation for high-quality mass-production.
We used 1x NVIDIA RTX™ A6000 (VRAM 48GB GDDR6 ECC) to generate images in this research showcase.
Machine-generated materials are not art or literature. We do not and will never consider our research as creative approach. This is only production approach.