This database* is an ongoing project to aggregate tools and resources for artists, engineers, curators & researchers interested in incorporating machine learning (ML) and other forms of artificial intelligence (AI) into their practice. Resources in the database come from our partners and network; tools cover a broad spectrum of possibilities presented by the current advances in ML like enabling users to generate images from their own data, create interactive artworks, draft texts or recognise objects. Most of the tools require some coding skills, however, we’ve noted ones that don’t. Beginners are encouraged to turn to RunwayML or entries tagged as courses.

*This database isn’t comprehensive—it's a growing collection of research commissioned & collected by the Creative AI Lab. The latest tools were selected by Luba Elliott. Check back for new entries.

Via : https://docs.google.com/document/d/1TkusCE5mS4tuTYoBwaTV4aJKdSVsf9jFKsGJCx8M05c/edit

Paint and Animate Live!

Tagtool transforms your iPad into a visual live instrument. Paint with light, create animated graffiti or tell improvised stories.

https://www.tagtool.org/
https://bibliogram.art/u/tagtoolcrew

3d collages composed in Unity with scans collected in botanic gardens, backyards, beaches, natural parks…

The shutdown of public life is accompanied by a new understanding of limitation. National borders have been closed, environments have shrunk to the nearest surroundings, and a lack of information and events makes it more difficult for new ideas to emerge. At the same time, the question arises as to whether creative potential is also hidden behind personal limitation. One of the first answers to this question is celebrating its opening this week.

SHUTDOWN.gallery is the latest project of Patrik Hübner. The German designer has set himself the goal of adding a new dimension to cultural life in times of social distancing.

Since 2018 Hübner has been pushing the idea of digital space and the concept of the online gallery. In those days a little ahead of its time, the project is now coming to life in the current events. "I had to watch as museums and galleries in my area closed their doors and the experience of art came to a standstill. It felt as if the gallery had to exist right now."

SHUTDOWN.gallery is a first step towards understanding the analogue and digital world together. It is intended to add a new level to the construct of what a gallery is and to blur the boundaries between analog and digital experience.

A logical next step in developing a narrative for digital spaces that can be experienced. References to real objects and the associated spatial limitations are important to facilitate the transition into this digital space and to prepare metaphors for future reinterpretations of the gallery concept.

The digital gallery adapts to the physical situation of the viewer and is freely accessible: On mobile devices such as telephones or tablets, the visitor can view the room directly and from all perspectives by tilting the device or moving it freely in the room - the boundaries between the physical and the digital disappear suddenly and intuitively.

As a web-based project, the gallery is publicly accessible worldwide. Every week there will be a new exhibition with works that move the world of art and design, thus maintaining the fertile environment of constant change. A growing list of world-renowned designers and artists has already confirmed their participation in what is the start of a new chapter for galleries.

The gallery’s dynamic and flexible visual identity reflects this constant change and desire for collaboration and new perspectives by putting every week’s exhibition at the center of the visual campaign. The result is highly unique mix of the gallery’s space and the artist’s work which can be used across social- and physical spaces to promote and visually explore this collaboration.

Manifest

Shutdown Gallery overwrites the walls of your private space and keeps your mind in motion even though the physical world is standing still.

Growth is a fundamental aspect of human nature. We are used to increasing the space around us, to crossing borders, to thinking of and discovering unexpected grounds.

In this fertile environment of constant change, shutdown confronts us with new levels of restriction and limitation. Country borders are closed, private life reduced to a minimum and cultural life locked in silence. The question "What's new?" is increasingly left unanswered and so are the contents that used to stimulate our thoughts and conversations.

In absence of new impressions, SHUTDOWN.gallery appears as a contemporary place of visual experience that intends to reawaken cultural life beyond social distancing and to literally overwrite the walls of our private space.

Based on the idea to use limitation as a creative tool instead of declaring it as constraint, SHUTDOWN.gallery aims to rethink existing concepts and reestablish focus in a world of distraction.

Every week, there will be a new exhibition with an artist whose work moves the world of art and design and thus maintains the fertile environment of constant change.

Animation playground

We’d like to show you a new way to create hand-drawn animation. So playful you’ll think it’s a toy, but it's secretly a robust app with a reimagined UI for animators and creators.

Wonderfully musical

Taking inspiration from music creation tools, using Looom feels more like playing an instrument than operating software - exploring lines, shapes and colors through loopable time and rhythm.

Everyone in the loop

Looom is all about looping animation with a smile. It was designed for professionals and beginners of any age, and is already being used by some influential indie animators.

https://apps.apple.com/app/looom/id1454153126?

https://www.instagram.com/explore/tags/madewithlooom/

Stream HD : http://stream.p-node.org/dab.mp3

∏-node is an experimental platform for the development of an hybrid web/FM/DAB+ radio format. Through the interlinking of different approaches, tools, technologies and networks a decentralised broadcast structure is established where each of the network’s nodes serves to both receive and transmit information. Such a structure seeks to break with the classic one-way communication scheme, substituting it with a horizontal peer-to-peer model.

∏-node wants to explore the many dimensions of radio – its physicality (ether, radio waves and the electromagnetic spectrum), its spatiality (bandwidth, frequencies), its infrastructure (network of radio receivers/emitters), its methods of production and editorial content management (programming boards/teams, recording studios), its methods of metadata reception (RDS/SDR), its history (free radio and pirate radio movements), and its legislation. Most importantly, π-node also wants to examine the future role and potential of radio in a time when everything is going digital.

This is an interactive editor for making face filters with WebGL.
The language below is called GLSL, you can edit it to change the effect.

Libre and modular OSC / MIDI controller :
https://github.com/jean-emmanuel/open-stage-control/releases

Raymond Queneau (1903-1976) was a French writer, closely associatd with the surrealists.

Cent mille milliards de poèmes is a book of ten poems. From them you can create 100,000,000,000,000 properly-formed sonnets.

Queneau was a founder of OULIPO, a group devoted to exploring and playing with language.

The book's exraordinary. Treat yourself. It's published by Gallimard and still in print.

This site is an English translation

You can see the basic poems or random poems, assemble your own poem or run a slide show.

New poems can be formed by replacing lines with any line that rhymes

Background material throws light on Queneau, the book and the translation. The poems are annotated.

http://www.growndodo.com/wordplay/oulipo/queneau.swf

https://archive.org/details/Microsoft_Windows_95_Video_Guide_1995_Uncut_VHS_x264_Aw.mkv

Pataphysics is perhaps best (mis)understood by looking at some of its recurring themes, such as exceptions, syzygies, anomalies, clinamen, antinomies, contradictions, equivalents and imaginaries.
Clinamen

The term clinamen was first coined by Lucretius when he needed to name the aleatory swerve of atoms in their descent described by Epicurus. Approximately two millennia later, Alfred Jarry resurrected this obscured concept as a key principle of pataphysics. Its influences can be found in the Situationists’ détournement, the Dadaists’ ready-mades and Oulipo’s verbal games, and so on. Lucretius had already linked the indeterministic property of the clinamen to free will and the Oulipo interprets it as a chance to escape certain restrictions given that any initial constraint are still followed (just as the atoms don’t randomly start to ascend but they swerve). Experimental poet Christian Bök has called the clinamen the smallest possible aberration that can make the greatest possible difference.

One good example of a clinamen in action is Jarry’s merdre (the very first word in his Ubu play). He squeezed an extra ‘r’ into the French word ‘merde’ (meaning shit) and translates into something like ‘pshit’. By rendering a useful (if rude) word useless in this way, he introduced a pataphysical sense of creativity that persists: the word still exerts a fascination today.
Syzygy

A syzygy both surprises and confuses. The concept originally comes from the field of astronomy where it denotes the alignment of three celestial bodies. In a pataphysical context it usually describes a conjunction of things, something unexpected and surprising. Unlike serendipity, a simple chance encounter, the syzygy has a more scientific purpose. A typical instance is the pun, which Jarry called the syzygy of words. Next to being intentionally funny, puns demonstrate a clever use (or abuse) of grammar, syntax, pronunciation and/or semantics, often taken to a quite scientific level, such that without understanding of what is said and what the intended meaning is, the humour of the pun might be lost.
Antinomy

The antinomy, in a pataphysical sense, is the mutually incompatible or paradox. Mutually contradictory opposites can and do co-exist in the pataphysical universe.
Alfred Jarry

Alfred Jarry (1873-1907) is without doubt the prime exponent of pataphysics. The word pataphysics was invented by him and some of his schoolmates in France in the 1880s and Jarry elaborated on that initial idea, both in his celebrated Ubu plays and in his novels and speculative writings. He has been described as a poet, dramatist, novelist, journalist, artist, eccentric, alcoholic and sometimes even a lunatic.
Technical details

In short, the tool reads in a library of plaintext files, and creates an index (a dictionary type data structure storing the vocabulary of the whole corpus together with a list that contains all documents and positions of the term within the document in the vocabulary). There are two collections of texts to choose from, either the Faustroll corpus or the Shakespeare corpus at this point.
Index:
{word1: [[fileA, posa], [fileB, posb], ...], word2: [[fileC, posc], [fileK, posk], ...], ... }

All texts in the corpus are read into memory and processed, for example any stopwords of the source language are removed.

Once a user submits a query, various important functions are triggered. First, the three patalgorithms are run to populate a list of results to be rendered.
Results:
[(title, (pre, word, post), algorithm), ...]

Each algorithm pataphysicalises the original query term in its own way and looks for matches in the index.

Results are presented in one of three ways. The default is the poetry view. It displays 14 lines of text, each of which can be changed to another iff more results are available. This is heavily inspried by Raymond Queneau's 'Cent mille milliards de poèmes'. The other two options show the results either sorted by their source or by the algorithm by which they were generated.

Extreme Mercator projection with Marseille (France) as reference point (on the right side).

The map you see is the Mercator projection.

But unlike a standard Mercator projection, you can substitute any point on earth as the "pole". (The initial view shows Boston as the pole point)

Furthermore, this map cuts off much, much closer to the poles than normal, allowing you to see many more orders of magnitude of distortion.

Because this yields a map several times taller than it is wide, it is shown sideways from its usual orientation.

Backstory

The Mercator projection is infamous for its distortion at high latitudes. This distortion gets exponentially worse as you approach the poles. It is in fact impossible to show the poles on a Mercator map — they are infinitely far away.

Any Mercator map you've ever seen must cut off the top/bottom edges at some arbitrary point. The map usually stops hundreds, if not thousands of miles short of the poles.

But I've often wondered what lies beyond those cut-offs... to make a map that didn't cut off but simply kept going. As the distortion progresses towards infinity, you would eventually reach the scale of cities, houses, insects, atoms...

But of course that'd all be on a featureless expanse of ice.

To make things actually interesting, we must artifically shift the pole of the projection to a more interesting place. Imagine the earth encased by a rigid cage of latitude and longitude lines. We rotate the earth while leaving the cage fixed until a new point of interest has taken the place of the North Pole.

This is called an oblique Mercator, and is normally used to shift an area of interest onto the equator of the map to avoid distortion. But whereas others avoid the distortion, we embrace it.

Note how strange the oblique Mercator looks even without the increased cutoffs. The standard Mercator is so ingrained in the public consciousness that we perceive it as 'normal'. But once you shift the pole its pervasive distortion is shockingly apparent.

Flux Æterna is a musical composition for Internet. It comes in the form of an endless audio stream. Listening conditions of Flux Æterna are similar to those of a web radio. But here, the user can influence the contents of this work by providing his own sound files.

This work can be adapted to multi-channel installation or concert.
Technical setup

​

The composition of Flux Æterna is a technical set including:

1 : The website to connect to the stream, to upload sound files to the sftp server and to find information about the work.

2 : The score, written in the form of a software made in MaxMSP. Therefore, a software that reads and converts the sounds arriving at the server, following differents features deterministics, randoms or stochastics. A transcription, as musical notation, is attached to this program.

3: The audio stream, is published on the Internet through a server broadcasting in format mp3 stereo 320 kbps. It can be read by many softwares : VLC Quicktime iTunes Flash... running on any platform (PC, smartphone or tablet).

Listen to the flux here : http://srv-aeterna.univ-st-etienne.fr:8000/fluxaeterna

The Organization for Transformative Works (OTW) is a nonprofit organization, established by fans in 2007, to serve the interests of fans by providing access to and preserving the history of fanworks and fan culture in its myriad forms. We believe that fanworks are transformative and that transformative works are legitimate.

We are proactive and innovative in protecting and defending our work from commercial exploitation and legal challenge. We preserve our fannish economy, values, and creative expression by protecting and nurturing our fellow fans, our work, our commentary, our history, and our identity while providing the broadest possible access to fannish activity for all fans.

This is a directory of 249 links in 73 categories.

This directory is somewhat inspired by the old, failed link collections like the original Yahoo! and DMOZ. They were terrible—you couldn’t find anything, but what you did find was often unexpected. My ‘archivist’/‘forager’ tendencies want to do this.

Linking has kind of died in the wild. Google views a site like this as a link farm—so, directories have died off. Yeah, well, I find many of the ‘link farms’ in my Web/Directory list to be immensely ‘great’ and ‘satisfying’. More than anything, I hope mine intrigues you to build your own. This directory forms my connection to the rest of society.

I reserve the right to link to dipshits and crazies. I link to what piques my curiosity, what amazes me or what horrifies me. This includes you. (You know you want to participate.)

You might also look at it like: maybe I’ve friended these links. But instead of putting them in a big number that represents my friends—my 249 friends, you see—I list my friends out neatly and try to coax you to meet them.

Perhaps there is no need for friending. For likes. For upvotes. For hashtags. For boosts. For trending. For rank. For followers. For an algorithm.

Perhaps plain ole linking—and spending time telling you why I linked—is good enough, was always good enough. Perhaps it’s superior!

"MindCandy Volume 1: PC Demos." The movie explores the definition and roots of an artistic computer sub-culture known as the demoscene, a community that's given me endless inspiration since the 1980s.

http://mindcandy.de/

A visualization of global weather conditions
forecast by supercomputers
updated every three hours

ocean surface current estimates
updated every five days

ocean surface temperatures and
anomaly from daily average (1981-2011)
updated daily

ocean waves
updated every three hours

aurora
updated every thirty minutes

ComboGAN: Unrestrained Scalability for Image Domain Translation Asha Anoosheh, Eirikur Augustsson, Radu Timofte, Luc van Gool In Arxiv, 2017.

https://arxiv.org/pdf/1712.06909.pdf

This year alone has seen unprecedented leaps in the area of learning-based image translation, namely CycleGAN, by Zhuet al. But experiments so far have been tailored to merely two domains at a time, and scaling them to more would re-quire an quadratic number of models to be trained. And with two-domain models taking days to train on current hardware,the number of domains quickly becomes limited by the time and resources required to process them. In this paper, we pro-pose a multi-component image translation model and training scheme which scales linearly - both in resource consumption and time required - with the number of domains. We demonstrate its capabilities on a dataset of paintings by 14different artists and on images of the four different seasons in the Alps. Note that 14 data groups would need(14choose2) =91 different CycleGAN models: a total of 182 genera-tor/discriminator pairs; whereas our model requires only 14generator/discriminator pairs

UNIT: UNsupervised Image-to-image Translation Networks : https://github.com/mingyuliutw/UNIT