Stability AI has released a new AI model, a stable virtual camera. It claims that 2D images can be converted into “immersive” videos with realistic depth and perspective.
Virtual cameras are a commonly used tool in digital filmmaking and 3D animation to capture and navigate scenes in real time. With a stable virtual camera, stability has sought to add generated AI to the mix to provide greater control and customizability, the company states in a blog post.
A stable virtual camera generates a “new view” of the scene from one or more images (up to 32 total) at user-specified camera angles. This model can generate videos that move along a “dynamic” camera path or preset, such as “spiral”, “dry zoom”, “moving”, “pan”.
The current version of Survey Preview, a stable virtual camera, can generate videos with up to 1,000 frames of aspect ratio length for squares (1:1), portraits (9:16), and landscapes (16:9). Stability warns that models can produce poor quality results in certain scenarios, especially in images that feature “dynamic textures” such as humans, animals or water.
He states that “very vague scenes, complex camera paths that intersect objects and surfaces, and irregularly shaped objects are especially when the target view is very different from the input image, especially when the target view is very different.”
Stable virtual cameras can be used in research under a non-commercial license. You can embrace the face of the AI Dev platform.
Stability, a plagued company behind the stable spread of popular image generation models, raised new cash last year as investors, including Eric Schmidt and Napster founder Sean Parker, tried to turn the business around. Stability co-founder and EX-CEO Emad Mostaque has mismanaged stability to financial ruin, causing staff to step down, leading a partnership with Canva, and investors are reportedly concerned about the company’s outlook.
Over the past few months, Stability has hired a new CEO, appointed “Titanic” director James Cameron to the board, and released several new image generation models. In early March, the company worked with chip maker Arm to provide an AI model that can generate audio, including sound effects on mobile devices running the arm chip.
Source link
