Google has introduced its advanced artificial intelligence (AI) text-to-image generator tool Imagen 3 publicly. The tool will be available with Google’s ImageFX platform.
The image generator can be accessed on Google’s AI Test Kitchen. Google first introduced Imagen 3 during its Google I/O event that took place in May. However, the tool is currently available in the United States (US) only.
Google rolls out Imagen 3
Google’s DeepMind team explained in its official blog that Imagen 3 is trained on a large dataset. This dataset can include texts, images and associated annotations. The team further explained that in order to maintain safety standards and high quality, they have added a multi-stage filtering process in the tool.
The company further claims that the latest Imagen 3 will be better than some of the current top listed AI text-to-image generators such as DALL-E 3, Midjourney, and X’s (previously Twitter) Grok-2, among others.
Decoding Imagen 3
So what is Imagen 3 and how is it different from any ‘text-to-image’ generator tool? Imagen 3 can be defined as a type of artificial intelligence that can generate images from text prompts/descriptions given by the users. The company further added that Imagen is a diffusion model which has a ‘high degree of photorealism’.
Furthermore, the text-to-image AI tool can also be used in several other application developments. These includes the following:
- The tool can be used to generate mockups
- You can also create prototypes using Imagen 3
- You can create illustrations of your favourite images
Additionally the text-to-image tool can test data, create educational content, and generate visualisations for debugging. Google highlighted that the key to its text-to-image model are Imagen, Muse and Parti. Other than generating images with prompt, it can also detect explicit or sexually abusive contents. For example in case someone prompts any commands related to such content, the tool will not be generating images. In any form of harmful query it will just decline the request.
The way ahead
Earlier this year in February, Google had disabled Gemini AI bots from generating images, after it faced complaints of producing ‘inaccurate images’ of people searched for. Many users complained that they were not getting correct images results for their search. The Gemini AI bot either produced wrong images or irrelevant images that were not connected to the queries.
Moreover, as of now the AI-text-image generator tool Imagen 3 is available for public testers in the US. It is expected that Google will bring the tool to other global regions including India in the coming months.
Follow FE Tech Bytes on Twitter, Instagram, LinkedIn, Facebook
