Google entered into most in-demand feature for Creators
Google has finally launched image generation through Google Bard, one of the most coveted features of Generative AI. Google has also launched its most powerful and advance Large Language Model Gemini Pro globally marking one more milestone in Generative AI adoption.
With the launch of image generation feature Google is well positioning itself to compete on equal footing with rival Open AI’s ChatGPT with integration of DALLE-E3.
Important Features with Image Generation and Gemini Pro:
1. Gemini Pro Widespread Availability:
In December 2023 Goole launched its most advanced model Gemini in English as discussed in our previous blog and on Feb 1st 2024, Gemini Pro has been made available in over 40 languages and more than 230 countries and territories making it one of the most widely available Large Language Model.
The LMSO Large Model Systems Organization, (Leading Evaluator of Language Models and Chatbots across Languages) recently reported that “Bard with Gemini Pro” is one of the best chatbots available (both free and paid) and “has made a stunning leap forward”. And in our blind evaluations with our 3rd party raters, we found Bard with Gemini Pro to be one of the best conversational AIs out there, compared to the top free and paid alternatives.
2. Free to Use Image Generation:
Through Bard, Google is planning to give tough competition to numerous text to image prompts including DALLE-E3 with free to use launch. This feature is available in English language in most countries.
3. Updated Imagen 2 Model:
Text to Image prompt in bard has been powered by an updated Imagen 2 model which is the same model used by developers and enterprises like Canva and Snap. It’s available through Imagen API on Google cloud Vertex AI already being used by millions of users for their creative tasks. In December 2023, Google introduced updated Imagen 2 model which delivers high-quality, photorealistic outputs. New Imagen 2 Model has been well integrated into Bard and ImageFX for widespread use for creators/ bloggers/ artists making image creation more prevalent and accessible.
Imagen 2 has been developed by Google’s DeepMind with latest advancements in diffusion based model. This update is particularly useful to create realistic giving real life feel to images though at times compromising on artistic and more creative side that creators are used to while applying DALLE E3 or Microsoft Designer both of which are powered by the same model.
4. Imagen 2 Availability on Bard, ImageFX, SGE and Vertex AI:
Google has integrated Imagen 2 across its products for accessible to individual users, creators and enterprises. Integration into Bard has made it simple and seamless for creators to get creative and realistic images with simple prompts with applications ranging from Brochures, Presentations to Profile Creations. Imagen 2 also delivers high quality image creation in SGE and enterprise customers can access Imagen 2 through with Google Cloud’s Vertex AI.
Google has launched ImageFX to foster creativity by allowing users to create an image with a text prompt and then modify it with a new version of the prompting using the expressive chips. ImageFX is part of the AI Test Kitchen where you can experiment and create with the other generative artificial intelligence experiments such as MusicFX or TextFX.
Safety and Limitations:
- Image generation in Bard is available in most countries, except in the European Economic Area (EEA), Switzerland, and the UK. It’s only available for English prompts.
- For now, this feature isn’t available to users under 18.
- Imagen 2 applies filters to avoid generation of images of named people.
- Google technical guardrails and safety training prohibits violent, offensive and sexually explicit contents.