Google Has repeatedly shown what its various mechanical algorithms can do, e.g. MUM Or LaMDAAnd continues to reflect them New model artificial intellect Entitled “Picture”.

According to Jeff Dean, head of the company’s artificial intelligence, it promises to “foster creativity between people and humans. Computers“Y Be able to generate images from a single, short and simple text description.

The “picture” is very similar DALL-E-2, Developed by artificial intelligence Open AI (The company was founded Elon Musk) Which also allows you to create an image from a written description. However, there are some differences between the two models, Such as the level of detail and efficiency in creating this image.

Concrete, Google claims that its AI delivers results with a much more accurate level of detail than other systems. To verify this, the company created a benchmark called Chair, Which compares your AI model with similar models and presents the results side by side for the user to see. The “human evaluator” can distinguish them and choose the most realistic.

These people, according to the company, concluded that The photos produced by ‘Imagen’ were of a higher quality and “the images fit the text described”, Better than other models.

Artificial intelligence is more effective than others

‘Image’, Google’s artificial intelligence that shapes an image from a brief description of the text, is also “more computationally efficient, more efficient in memory and faster to fold” thanks to its proprietary architecture. AT. ორAs a result, Surreal images are made more accurately than any other model And description of any type of text.

Transparent duck sculpture made of glass. The statue is in front of a landscape painting, “Several robots having dinner against the backdrop of the Eiffel Tower,” or “A dog curiously looks in the mirror, sees a cat.” These are phrases that Google uses as an example to show what its AI model can do.

These are just some of the goal setting shareware that you can use Website:

Google, on the other hand, claims that “images” can also generate images based on descriptions of specific locations or even complex text. For example, if a user logs in “Giraffe runs with a rabbit at La Bombonera Stadium in Buenos Aires” The company AI will generate a picture based on this description and understand the scientific name of both the animal and the places.

The “image” is, at this point, Internal project and inaccessible to the public, Because it can lead to the creation of images that contain “Stereotypes and harmful perceptions”, As explained in the company.

“The image relies on text encoders, which are based on unverified web-wide data, and thus inherit the social biases and limitations of large linguistic models. As such, there is a risk that Imagen has encrypted harmful stereotypes and perceptions that guide our decision not to release Imagen for public use without additional warranties. Mention Google.

How to see what Google Account data apps are accessing

Jan. mobile phone:

1. Open Gmail.

2. Press the three horizontal stripes On the upper edge.

3. Go To install.

4. Click on the configuration account (if more than one is added to the mobile).

5. Click Manage your Google Account.

6. Scroll to the right setting safety.

7. Finally enter the section Third-party apps that can access your account.

– Jan. Web Navigator:

1. Press the button on the computer Dotted square Which appears next to the username on the top right edge.

2. Enter Bill.

3. Click on the right edge where it says safety.

4. Finally press the button Third party applications Who can access your account.

Keep reading