34279Automated Texture Generation for E-commerceClosedCollaborative R&DInnovate UKThere has been a surge in demand for 3D assets mainly due to emerging technologies such as Virtual Reality, Augmented Reality, 4K games, robotics, and photo-realistic rendered images for e-commerce. However, 3D asset generation remains time consuming and completely manual. This has limited the use of 3D assets to high budget movies and advertisements. ZEG.ai has developed a world-first 3D AI that allows for massive generation of 3D assets using intuitive approaches: 1) uploading an image 2) textual or voice descriptions, or 3) inference from spatial context. This enables faster 3D asset generation, enabling anyone to build and use 3D models through a simple web API. 3D modelling of any physical item comprises shape and texture. ZEG.ai has built the technology that allows you to rapidly build the 3D shape. However, creating and applying textures to a 3D shape is still a manual and slow process. In order to tackle this challenge, ZEG.ai has partnered with Cyanapse to create an AI tool that allows for the automatic generation of textures using a single photograph. Cyanapse will be leading the machine learning research aspects of the project and will work together with the ZEG.ai team to create a proprietary tech that is immediately usable in the 3D workflow. The key technical innovation of this project involves the development and validation of new generative deep learning models for texture synthesis for realistic 3D object creation in the context of e-commerce items. The ability to generate high-quality and realistic 3D object building with such ease will have a revolutionary impact on industries that require 3D graphics modelling and design.