Issue #4 July 10th - August 15th

🤖 The Creative AI newsletter 🎨

by Luba Elliott

Hi <<First Name>>! You've been to one of my Creative AI meetup events and have opted in to receive occasional emails with updates on creative applications of artificial intelligence in art, music, design and beyond. Here's issue #4 😀

🎵 Anime, Music, Fashion 👗

Alejandro Giacometti from EDITD used deep learning techniques on a dataset of 3,000 images from London Fashion Week to detect trends, brand identities of individual designers and more. Meanwhile, University of Maryland researchers wrote a paper on using style transfer to design clothes based on user preferences.

Now you can also generate customised new anime character designs on the MakeGirls.moe website based on a GAN model developed by researchers from Fudan University, Carnegie Mellon University, Tongji University and Stony Brook University.

The Guardian published an overview of AI music start-ups, including AI music, who are working on a tool to enable songs to shift themselves depending on the context they are being listened to. There's also been some research from Birmingham City University on an AI engine that can predict how musical legends would play contemporary music.

🎨 Art 🎨

Anna Ridler gave an interview to Alphr about her practice incorporating life drawing and machine learning with a focus on her recent works Fall of the House of Usher and Drawing Sound. Read on here.

Robbie Barrat pre-trained DCGAN models on art datasets including landscapes and portraits with some interesting results when resuming from a checkpoint and training on different data. Learn more here.

Sascha Pohflepp has released a new work called Spacewalk based on a customised GAN trained on 3,000 images of predatory animals to see what patterns emerge. For some hazy images of jaguars and the artist's intention behind the work, go here.

Machine learning as inspiration for physical paintings. Leonard Bogdonoff used style transfer to combine graffiti and waves and then paint the resulting image.

Computed Curation is a photobook created using t-SNE, genetic algorithms and computer vision techniques from an archive of pictures. See it on Philipp Schmitt's website.

📓 Resources 📓

The Center for Open Data in the Humanities released datasets of pre-modern Japanese texts and character shapes, Edo cooking recipes and modern magazines from 1880s. Get them here, though you may need the help of Google Translate.

Proving Ground have released new machine learning tools for Grasshopper called LunchBoxML, which enables users to work with regression algorithms, clustering models, and neural networks within the computational design workflow.

There's been a lot of relevant workshops recently, so if you have time check out Machine Learning for Creativity workshop (homepage), Procedural Content Generation workshop ( livestream I homepage) and Machine Learning for Music Discovery workshop (homepage I review).

You can find the recordings from the two days of the CFI conference themed on AI: Myth and Reality and Trust in the Age of Intelligent Machines here and here. My talk on Art + AI is on at (2:40:40) on the first day.

🚀 Opportunities 🚀

Bernstein Network Computational Neuroscience has an open call for artists interested in the philosophical, ethical and socio-cultural aspects of computational neuroscience. The selected artist will produce a work to be exhibited at the Bernstein Conference 2018 in Berlin. Apply before 31st August.

Erica Scourti is looking for a technologist (ideally London-based!) to collaborate with for her next artwork. The work includes building a chatbot based on previous social media messages and working with an emotional valence API to translate voice recordings into visual and/or sonic outputs. Ping me if you are interested and I will share more details.

PROCJAM are taking applications for their first ever £1000 grant to help support someone working with generative software. Details on how to apply before 31st August.

🤖The Meetup 🎨

The next event will be on the 30th August with Emily Denton and Lucas Theis talking about their work on learning image representations. Registration will open at 10am on the 16th August. Follow the meetup group for updates.

Slides from the July event on Calligraphic Traces are now online. You can find Daniel Berio's presentation on Graffiti synthesis, a motion centric approach here and Georgia Ward Dyer's slides and notes from her talk O time thy pyramids here . There's also a periscope recording.

😃Cool things to do 😃

The QUAD in Derby has an exhibition on Machine Made: Art, Robots and AI featuring our friends Jake Elwes, Marisa Tapper and Rob Homewood. Drop by if you're in the area before 10th September.

Saatchi Gallery's exhibition From Selfie to Self-Expression features some cool interactive artworks incorporating facial recognition and computer vision techniques including Zoom Pavilion by Rafael Lozano-Hemmer and PomPom Mirror by Daniel Rozin. On until 6th September.

The Ars Electronica festival of media art is AI-themed this year, focussing on "The Other I - Technology as Antagonist or Alter Ego?" between 7th-11th September in Linz. The programme includes exhibitions, conferences and a hackathon. See highlights here.

Thanks for getting this far. Anything I missed? Drop me a line if you have any updates I should include or if you need any creative AI-related help.