DALL-E and other AI tools that can create fantastic images have drawn attention to generative AI models. But their capabilities are not limited to this – artificial intelligence already knows how to create texts, invent fonts and depict our speech in pictures.
Want to use AI in your art? Get started with these apps, programs, and tools. Some of them can be used even without experience.
The best visual creation tools
- Magenta is an open-source research project exploring the role of machine learning as a tool in the creative process. Requires programming skills.
- Processing is a flexible note-taking program that teaches programming in the context of the visual arts. Includes p5js (for JavaScript) and Processing.py (for Python). Does not use AI, but is suitable for working with generative visual art.
- ml5.js is a tool that aims to make machine learning accessible to a wide audience of artists, creative developers, and learners.
- DeepArt.io – allows you to upload a photo and apply various artistic styles to it.
Visionist is a tool that edits images using AI-based styles to create abstractions, contour portraits, and more (iOS version is Deep Art generator developed by 3DTOPO Inc.).
- GoArt – AI effects make photos look like famous portraits (Web, Android and iOS versions available. Developed by Fotor).
- Deep Angel – The editor automatically removes objects or people from images.
- Fractal Art Generators – do not use AI as such, but are no less interesting because they belong to the field of art generated by a machine.
Image processing and creation
- AutoDraw – Turns a sketch into a graphic image.
- AI Painter – turns photos into AI paintings or creates abstract images using a neural network.
- Quick, Draw! – a game in which the neural network tries to guess what the user is drawing.
- Sketch-RNN Demos – draw with a neural network.
- Cartoonify – Turn your portrait into a computer generated cartoon image.
Words and texts
- GPT-2 is a computer language model created by OpenAI. Generates synthetic text if given a start.
- Handwriting with a neural net – play with a neural network that creates handwriting depending on your writing style.
Sounds and music
- Magenta Studio is a collection of music plugins based on open-source Magenta tools and models.
- AI Duet is an interactive piano.
- NSynth Sound Maker – allows the user to compose their own hybrid sounds and instruments.
- MuseNet – The tool plays 4-minute musical compositions with 10 instruments and can combine styles from country to Mozart using MuseNet (also available on GitHub).
Movement and dance
- AI Sketches With Bill T. Jones – The PoseNet model literally learns “body language” by assessing the relationship between body position, speech, and dancer movement.
Voice and interpretation
- Scribbling Speech – Turns speech into animated drawings in real time.
- Thing Translator – reproduces the name of an item in different languages, using only its photo.
Visualization of neural networks
- Activation Atlases – creates an atlas of functions learned by the neural network.
- What neural see networks – allows you to connect a camera and evaluate how the neural network perceives the image. Equally interesting is the Feature Visualization tool .
- Visualizing high dimensional space – visualizes how machine learning works.
- t-SNE – Creates 2D data maps with hundreds or even thousands of measurements.
Training of neural networks
- Teachable machine – allows you to train a neural network in real time and right in the browser using your own camera. Doesn’t require you to write code.
Typography
- FontJoy – Creates font combinations using deep learning.
- Font Map – finds new relationships between fonts using machine learning.
Generative Adversarial Networks
GAN (Generative adversarial network, generative adversarial networks) – machine learning algorithms of this kind are built on a combination of two neural networks, one of which creates samples, and the other seeks to distinguish genuine samples from imitations. Often used to generate realistic images.
- GAN Zoo is a list of Generative Adversarial Networks (GANs) that is updated from time to time.
- GANS Awesome Applications – A selection of neural networks and GAN demos.
Libraries for machine learning
- TensorFlow.js is a library for developing and training ML models in JavaScript.
- scikit-learn is one of the most popular data collection and analysis tools. It is based on NumPy, SciPy and matplotlib and is open source, suitable for commercial use (BSD license).
