You can now generate photos faster with Getty Images’ improved AI model
Getty Images on Tuesday upgraded the artificial intelligence (AI) image generation model that it integrates into its platform. The updated AI model is now built on the Nvidia Edify model architecture and offers faster generation speeds and a greater number of variations on a single prompt. The company also claims that the new image generator adds more detail to 4K images and accurately follows prompts, even when they are long. The AI model will power the company’s AI tools, including Getty Images’ Generative AI and iStock’s Generative AI.
Getty Images updates AI model
In a newsroom afterthe company announced the updated AI model. It has now been released and is available for testing on the Getty Images platform. The new image generation model is built using Nvidia’s Edify architecture, which is part of Nvidia Picasso, an AI foundry where software developers can build and deploy generative AI models for visual design.
The company highlighted that the Edify architecture brings significant upgrades to the model, including higher output speed and quality, better alignment with prompts, and improved 4K upscaling. With this, the AI model can now generate four images in six seconds, which the company claims is double the previous speed.
Furthermore, prompt compliance is said to have increased with the new model. Getty Images says the AI can now support prompts of up to 250 words and handle complex requirements. Image modification capabilities are also available with the new large language model (LLM). Users can now add or modify individual elements, expand the canvas or easily remove backgrounds.
About training data, Getty Images said“The model was trained only on high-resolution images licensed or owned by Getty Images and metadata from Getty Images’ vast creative library.” The company also claimed that no data or images were scraped from the internet, synthetically generated, or derived from the output of other generators.
Additionally, the company claimed that the image generator is trained not to produce output that could violate intellectual property rights or artists’ rights. As a result, prompts requesting images of identifiable people, protected locations, trademarks or brands will not be executed.