Sothink - make a logo by hand

For those who find neural network logos too monotonous or just like old-school methods. Here you can create your own logo without any newfangled neural networks.

Simple templates, color tools - something that is a thing of the past, but can still come in handy if the AI is not able to do what is required. Works on both Windows and macOS.

#service

Meta strikes the next blow to the industry: they will soon release their model for working with code

It is called Code Llama and should be released next week. And as we all understand, the model promises to be free and open (hold on, Copilot and GPT4).

We are waiting!

#service

MatchColors - copy the processing of another photo

An application that allows you to colorize any photo or video by completely copying color correction and other effects from the reference you need. Just select your file and a suitable example - the neural network will do it all by itself.

Professional tools for color correction and fine-tuning are added just in case. There is a free version (with limitations), but a pro version is also available - from $1 per week.

#service

Here is a powerful compilation of the best AI add-ons for Stable Diffusion XL.

The catalog includes 3D Render Style, Watercolor Style, Cyborg Style. Ugly Sonic and other addons.

You can try it for free here: https://huggingface.co/spaces/multimodalart/LoraTheExplorer.

#service #useful

A new addition from Adobe is the Adobe Express neuronka. The service is powered by Adobe Firefly and is able to generate logo designs, ticktocks, resumes, rils and more. There are also thousands of free professional templates.

The paid version costs 9.99 bucks and offers full AI functionality.

#service #useful

Scientists have created a fully automatic AI janitorial machine called Somatic.

The robot cleans and disinfects any surface and works several times faster than a regular maid.

Cybercleaner services can be ordered for 1k dollars a month - and the robot will work 40 hours a week.

#news

Here are some useful AIs for working with Python. There's AI for APIs, machine learning, integrations, and UI.

Save them for yourself so you don't forget them.

#useful #guide

An enthusiast "revived" the covers of legendary vinyls.

Created with the help of Runway Gen2 and Pika Labs.

#aigeneration

🤖 New robots in the world of agriculture!

Introducing the Laserweeder3000 - an autonomous robot Laserweeder from an American company.
This advanced robot is equipped with a laser and equipped with artificial intelligence, thanks to which it is able to effectively destroy weeds🔥

The cost of this innovation is $1.2 million, but the manufacturer assures that the payback will come in 1.5-2 years.

This means that in the long term, investing in the weed killer 3000 is very profitable💰

What are the pros of this robot? 🤖

Modern chemicals are gradually losing their effectiveness due to weeds adapting to their effects🌱

Weedkiller3000 handles the task of weed control more effectively and safely.

#news #funny

An open source neural network for video processing - CoDeF.

CoDeF is able to translate image to image and video to video without training, as well as translate keypoint detection to keypoint tracking.

It requires a minimum of 10 GB of video memory.

Project repo 👈

#service #video

Imagine that in a text promt you write the word house in orange and Stable Diffusion changes the color of the house in the picture. You change the font and the neural network changes the object. More details on the attached video.

Also if you increase the font size, this object will have more weight.

Repo 👈 of the project with the guide.

#service

Midjourney is finally launching an inpaint option.

The new feature is called Vary and the button is located above the Zoom Out buttons.

Remix must be enabled. Then if you click on Vary, a new editor window will appear where you can select a part of the image and specify what you want to generate there.

Developers say that the new feature works better on large areas of the image (from 20% to 50% of the picture).

#service

Most likely you've seen neural network-generated horror instead of normal captions on the pictures.

The Textify 👈 neural network changes gibberish into normal text. Just upload your picture and write the text you want.

#service

But here is a new AI from former Google employees that generates text immediately as it should.

Ideogram is a new generative neural network for creating pictures. You can sign up quickly by logging in with a Google account. The project page displays both new pictures and those with more likes.

#service

The world's first robot pilot has been unveiled in South Korea. "Pibot" is AI-based and is equally at ease with both military and civilian aircraft. The cyborg independently climbs into the cockpit, switches toggle switches and freely memorizes flight map routes and emergency protocols.

Pilots are getting nervous...

#news

Video made using only AI and a plastic figurine

#aigeneration #video

Text > animated 3D avatars.

The TADA neural network (Text to Animatable Digital Avatars
) uses textual queries to generate 3D models with high-quality geometry and realistic textures, which can then be animated and rendered in traditional graphics programs.

The developers say that their model generates high-quality geometry, textures and facial features. The license is available for development.

Project website 👈

#service

The coolest browser plugin that combines 10+ top neurons like Claude2, ChatGPT and Bing.

It works simply - choose up to 5 desired AIs, write a text promt to the chatbot and get 5 different answers. Considering that each service is good in its own way - your work will be much more efficient.

Works on any top OS. See the link 👈

#service #useful

Microsoft has revealed an update to the iconic Paint for Windows 11. The graphics editor will add AI like Image Creator, which will allow you to create drawings via text command or draft.

Along with this, developers plan to update other OS services: in "Photos" will be able to find people from a photo, and through "Scissors" - to copy text from a fragment of the screen.

#news

⛔️💣 YouTube is against using AI-generated music in your content

The details are still hidden under a veil of secrecy, but some interesting speculation has already emerged. It is said that YouTube may introduce an innovative system similar to the famous Content ID.

The platform has already shared the key principles of the new initiative. They include responsible use of AI technology, ensuring content safety and copyright protection. Three key principles that have brought the music world together in harmony.

YouTube is also preparing an extensive update to its policies on AI-generated content. AI will also be involved here - to detect content that doesn't comply with the rules. But that's not all! A unique program similar to Content ID is planned, but this time for AI-generated music and creative partners. Authors will be able to monitor their work more closely.

YouTube is building a partnership with music giant Universal Music and its artists. The platform has even announced an "AI music incubator" that will help bring songwriters and producers together. What will happen next? Stay tuned for more updates!

🤔 Interesting development, isn't it? In April, Universal Music called on other platforms to remove AI-created music. Now Google and Universal Music are actively discussing licensing the voices and melodies of musicians. This opens up new avenues for AI-enabled creativity.

#news