Imagen 2, Google Cloud's latest text-to-image technology, is now generally available for Vertex AI customers.
Customization and Deployment
Vertex AI customers on the allowlist can customize and deploy Imagen 2 using intuitive tooling, fully-managed infrastructure, and built-in privacy and safety features.
Key Features of Imagen 2
Imagen 2, developed with Google DeepMind technology, excels in generating high-quality, photorealistic, and high-resolution images from natural language prompts.
Vertex AI's Indemnification Commitment
Vertex AI's indemnification commitment now covers Imagen on Vertex AI, providing customers peace of mind.
Enterprise-Ready Capabilities
Imagen 2 on Vertex AI caters to enterprise needs, offering image generation features with enterprise-grade reliability and governance.
New Features in Imagen 2
Highlighting the new features, Imagen 2 achieves high-quality photorealistic outputs, addresses challenges in text rendering, and introduces logo generation.
Multi-Language Support
Imagen 2 supports six additional languages beyond English and allows translation between prompt and output. This includes languages such as Chinese, Hindi, Japanese, Korean.
Safety Measures
Imagen 2 integrates built-in safety precautions aligned with Google's Responsible AI principles. It includes an experimental digital watermarking service.
Use Cases and Customer Testimonials
Snap, Shutterstock, and Canva showcase successful implementations of Imagen. Snap uses it for AI Camera Mode, Shutterstock enhances its AI image generator.
Getting Started with Imagen 2
Organizations can explore Imagen 2 on Vertex AI by referring to documentation or joining the Trusted Tester Program through their Google Cloud account representative.