Post Thumbnail

Stability AI introduces transformation of 2D photos into 3D scenes

Stability AI has announced the release of an innovative AI model called Stable Virtual Camera, capable of transforming two-dimensional images into immersive video materials with realistic depth and perspective. This technological achievement marks significant progress in the field of generative artificial intelligence and virtual cinematography.

Stable Virtual Camera represents a substantial development of the virtual camera concept, traditionally used in digital filmmaking and 3D animation. The integration of generative AI into this toolkit opens unprecedented possibilities for control and customization of created content.

The technical capabilities of the new model are impressive:

  • Generation of “new angles” of a scene based on one to 32 source images.
  • Creation of videos with dynamic camera movement trajectories.
  • Support for preset shooting modes: “Spiral”, “Dolly Zoom”, “Move”, and “Pan”.
  • Ability to work with various formats: square (1:1), portrait (9:16), and landscape (16:9).
  • Capability to generate video sequences up to 1000 frames long.

Despite the impressive capabilities, Stability warns about potential limitations of the current research version. The model may show quality degradation when working with images containing people, animals, or “dynamic textures” like water. “Highly ambitious scenes, complex camera trajectories crossing objects or surfaces, and irregularly shaped objects can cause a flickering effect, especially when target angles differ significantly from the source images,” the company notes in its blog.

This technological breakthrough comes during a difficult period for Stability AI. The company, known for its popular image generation model Stable Diffusion, attracted new investments last year from significant figures such as Eric Schmidt and Napster founder Sean Parker, who are seeking to restore the business. Earlier it was reported that Stability co-founder and former CEO Emad Mostaque led the company to financial problems, resulting in employee layoffs, the failure of a partnership with Canva, and growing investor concerns about the company’s prospects.

Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.

Latest News

How to create an infinite universe with one text prompt

Forget everything you knew about creating game worlds. Tencent just released the open-source model Hunyuan-GameCraft. Which generates interactive virtual worlds directly on your graphics card. Link in description. One text prompt — and you have an infinite universe.

How synchronization of 3 light sources protects against forgeries

Artificial intelligence has learned to create video fakes that are impossible to distinguish from reality. And this is a huge problem and question of trust in society. But scientists from Cornell University found a brilliant solution. They hid watermarks right in ordinary lighting.

Hip-hop, wushu and Peking opera at the robotics games opening ceremony

China hosted the first World Humanoid Robot Games where 280 teams from 16 countries competed. Who brought more than 500 androids. It became almost a real Olympics for robots with all the attributes of major sports.

The first LAARMA system protects animals on Australian roads

In Australia, animal-vehicle collisions are a serious problem for this continent's ecosystem. Now scientists have found a technological solution. The world's first roadside LAARMA system based on artificial intelligence that protects wild animals from dangerous encounters with traffic.

Nvidia introduced Cosmos model family for robotics

Nvidia company introduced the Cosmos family of AI models. Which can fundamentally change the approach to creating robots and physical AI agents.