Google released its in-house artificial intelligence (AI) model for image generation, Imagen 3, on Thursday. The tech giant did not make any announcement for the release, and instead released the model quietly to users. Currently, the text-to-image generation model is only available to users in the US. However, there is no word on when it will roll out to users in other regions.
Google releases Imagen 3 AI image generation model
The tech giant’s AI Test Kitchen is now allowing users to sign up to the platform. They can then use the AI model to generate images. The third generation of its Imagen model will get an improved texture generation and word recognition capabilities. Besides, there will be stricter prompt adherence. However, a Reddit user claimed that he was able to generate images in various styles. This includes Nikon DSLR quality, GoPro style, wide angle lens, and more. However, the model is struggling with generating close-up images with multiple people and underlit images.
Get to know about its functioning
The user also claimed that the model was producing erroneous results. Especially, when it was using prompts such as “a guy holding a cup of coffee”. The AI features are generating extra limbs, creating a random limb holding the object and more. The company highlighted that it used a latent diffusion model. This is a variant of the diffusion model popularized by Stable Diffusion. The company also added that it is using new methods to minimize the potential harm using the Imagen 3 model.
The free tier of the Gemini chatbot can also generate images, but it uses Gemini’s capabilities for this. Google has built Imagen 3 on a different architecture. Since its dataset largely contains images, it is better trained to generate AI images.
For more information please keep reading techinnews
