Lumiere

Video generation

Turning text into stylized videos.

Lumiere screenshot

About Lumiere

Introducing Lumiere, a revolutionary space-time diffusion model developed by Google Research, specifically designed for cutting-edge video generation. Lumiere excels in creating videos that showcase realistic, diverse, and coherent motion.With three key functionalities - Text-to-Video, Image-to-Video, and Stylized Generation - Lumiere offers a versatile tool for video creation. The Text-to-Video feature allows Lumiere to generate videos based on text inputs, providing a dynamic interpretation of the given prompts.Similarly, the Image-to-Video feature utilizes input images as a starting point for video generation. Lumiere's Stylized Generation capability adds unique styles to generated videos, using a single reference image to infuse the target style.Lumiere stands out with its Space-Time U-Net architecture, enabling the generation of entire videos in one seamless pass. This approach contrasts with traditional video models, eliminating the need for temporal super-resolution and ensuring the video's temporal consistency.From animals to nature scenes, objects, and people in various settings, Lumiere's applications span across different scenes and subjects, often presenting them in imaginative or fantastical scenarios. Its potential applications range from entertainment and gaming to virtual reality, advertising, and any other context requiring dynamic and engaging visual content.

Companies Are Making AI Skills Mandatory

Performance reviews and hiring now depend on AI proficiency

Meta
Shopify
Microsoft
Duolingo
Klarna
Google
Opendoor
Thomson Reuters
Fiverr
Amazon

Track the Impact of Your AI Usage

Document your productivity gains and build your AI portfolio for performance reviews

Start Tracking Free