A web application with multi-input capabilities for AI-driven 3D object generation, designed for XR applications

The generation of 3D content is essential in many applications and especially the ones relying on eXtended Reality (XR). However, creating XR-ready models is typically a costly and timely manual process, as well as a complex task for non-expert users. Here, a platform for multi-input 3D model generation is presented. It integrates, in a single point, methods to generate XR-ready 3D objects from three input types: multiple images, a single image, or a text prompt. Each input is managed by a separate service, offered to users either via an API, to support integration with other services, or via a dedicated intuitive web-application, designed for non-experts. The web application contains tools that ease the usage of developed services and allow scheduling operations, viewing results, and post-processing of 3D models in a user-friendly manner. It also provides additional features, such as simplification, remeshing, and levels of detail generation, developed to make the generated models even more suitable for XR applications.