Image to 3D

Trellis

A novel 3D generation method is introduced, enabling versatile and high-quality 3D asset creation. At its core is a unified Structured LATent (SLAT) representation, which supports decoding into multiple output formats, including Radiance Fields, 3D Gaussians, and meshes. This is achieved by combining a sparsely-populated 3D grid with dense multiview visual features extracted from a robust vision foundation model, effectively capturing both structural (geometry) and textural (appearance) information while preserving decoding flexibility.

Structured 3D Latents for Scalable and Versatile 3D Generation (https://trellis3d.github.io/)

The method utilizes rectified flow transformers specifically designed for SLAT as 3D generation models. These models, trained on a large dataset of 500K diverse 3D objects and scaled up to 2 billion parameters, demonstrate superior performance in generating high-quality results under text or image conditioning, outperforming existing methods, including recent large-scale approaches. Additionally, the method enables flexible output format selection and local 3D editing capabilities, features not supported by prior models. Code, models, and data will be made publicly available.

3D model and mesh

This allows to check the meshing process and quality.

Test in Twinmotion

 

Tripo

https://www.tripo3d.ai/app/home

Tripo AI is an innovative platform that leverages artificial intelligence to streamline the creation of high-quality 3D models. Users can generate detailed 3D models in seconds by inputting text descriptions or uploading images. This tool caters to a wide range of professionals, including designers, developers, and creatives, by simplifying the traditionally complex process of 3D modeling. Tripo AI’s applications span various industries, such as gaming, animation, product design, and architecture, making 3D content creation more accessible and efficient.

Creating a 3D model

Krea

https://www.krea.ai/

Krea.ai is an innovative platform that leverages artificial intelligence to facilitate the creation and enhancement of visual content, including images and videos. It offers a suite of tools that enable users to generate high-quality visuals in real-time from text descriptions or by modifying existing visuals. Key features include real-time AI image generation, enhancement and upscaling of images and videos, AI-driven video generation from text descriptions, and specialized mini-apps for creating logo illusions and patterns. Krea.ai caters to a diverse audience, including artists, designers, marketers, and content creators, providing an accessible solution for visual content creation and enhancement.

Creating 3D

https://www.krea.ai/apps/image/realtime

Using advanced AI models, it analyzes the image to create a depth map, allowing users to visualize and manipulate the scene with parallax effects. This feature is ideal for adding dynamic depth to 2D images, making it useful for animation, AR/VR applications, and cinematic effects. Users can upload an image and instantly see it transformed into a 3D-like scene with adjustable depth settings.

A 3D object can be freely moved in all three dimensions (X, Y, Z), making it easy to place it exactly where needed in a scene. In contrast, a 2D image is fixed and requires manual distortion to simulate depth.