The future of 3D modeling will be as effortless as hitting 'Print'

Stabilities' new model, SVD3, can create detailed 3D meshes from a single image of an object.

Based on Stable Video Diffusion, SV3D specializes in creating two distinct types of videos:

•⁠ ⁠Orbital videos with SV3D_u, showcasing various angles around an object

•⁠ ⁠Path-specific videos with SV3D_p for more directed visual narratives.

These videos are then employed for mesh creation, outputting detailed 3D objects.

Typically, 3D models are linked to video games and creating virtual environments.

However, transforming a single image into a digital format can unlock new avenues of use.

This allows e-commerce merchants to offer a richer experience of their products while educators can make subjects more engaging.

Utility is in the eye of the beholder, willing to see past traditional uses.

Previous
Previous

Anthropic researchers released a new way to bypass LLMs

Next
Next

If you can understand spreadsheets, you can understand how AI works