Stability AI Unveils Innovative SV3D Model for 3D Video Generation

Stability AI has just released Stable Video 3D (SV3D), a generative model that advances the field of 3D technology by improving quality and viewpoint consistency. Built on the foundation of Stable Video Diffusion, SV3D offers superior quality and multi-viewpoint consistency compared to previous and open-source alternatives.

The release of SV3D includes two versions: SV3D_u and SV3D_p. SV3D_u can generate trajectory videos from a single image without the need for camera adjustments. SV3D_p goes a step further by supporting both single image and trajectory viewpoints, allowing for the creation of 3D videos along specified camera paths. This flexibility and control open up a world of possibilities for content creators and developers.

One key advantage of SV3D is its use of a video diffusion model. Unlike the image diffusion model used in Stable Zero123, video diffusion provides significant improvements in output generalization and viewpoint consistency.

Internally, SV3D utilizes its multi-viewpoint consistency to optimize 3D Neural Radiance Fields (NeRF) and grid representations. This optimization is further enhanced through masked score distillation sampling loss, which aims to improve the 3D quality in unseen regions of predicted viewpoints. To address the issue of intrinsic lighting, SV3D adopts a separate lighting model that is jointly optimized with 3D shape and texture.

SV3D has a wide range of potential applications. From virtual and augmented reality experiences to product visualization, this technology has the ability to transform multiple industries. As Stability AI continues to push the boundaries of AI-driven 3D generation technology, it will be exciting to see how creators and businesses leverage the capabilities of SV3D.

Stable Video 3D is now available for commercial use through Stability AI membership. For non-commercial purposes, model weights can be downloaded from Hugging Face. If you are interested in delving into the technical details, the research paper can be found here.

With the release of Stable Video 3D, Stability AI once again demonstrates its commitment to driving AI advancements and providing cutting-edge tools for creators. As the company continues to innovate, it is evident that the future of 3D generation technology lies in the hands of capable individuals.