| Maymuunah |
https://www.artaigallery.com/ |
| Neela |
https://www.youtube.com/watch?v=PCBTZh41Ris |
| Daniel Torres |
https://www.artbreeder.com/ |
| kobe |
https://experiments.withgoogle.com/ai/beat-blender/view/ |
| Solomon |
http://wpmemo.memo.tv/works/learning-to-see/ Some super cool videos here! And would definitely recommend reading the text describing the project on the side of the page. |
| KChu |
http://qosmo.jp/projects/ai-dj-human-dj-b2b/ |
| Belinda |
It's called "What I saw before the darkness" under Visual, but that link isn't working, so here is a link to an article that has a video of the time lapse for that project: https://futurism.com/the-byte/watch-ai-die-neural-network |
| Belinda |
I also found a version of Generated Recipes; it's pretty amusing: https://trekhleb.dev/machine-learning-experiments/#/experiments/RecipeGenerationRNN |
| Jt |
https://www.memo.tv/portfolio/dirty-data/ was quite interesting to me because of its nature of using "dirty data", or not cleaning any input data used. In the examples they gave, the pictures have some really messed up faces that aren't exactly human-like, which reiterates the importance of the data that is used in training any machine learning model. |
| gwitchel |
I think my favorite is https://experiments.withgoogle.com/ai/beat-blender/view/ because of how well built out the idea is. They don't add any fluff, but the core idea is so amazing and the applications are very clear without being stated outright. |
| armwong |
My favourite is http://www.memo.tv/works/learning-to-see/ Since they were able to apply the technique to a live camera feed, it really showcases how efficient these algorithms can be. I also really enjoyed Gene Kogan's work, particularly https://genekogan.com/works/a-book-from-the-sky/ Although we saw a similar technique with photos, it's interesting to see the technique be applied to text. |
| alexwerner (she/her) |
I love the book 'Invisible Cities' by Italo Calvino and I was really excited to see that Gene Kogan had used it as a basis for a project using maps. https://opendot.github.io/ml4a-invisible-cities/ |
|
The Uncanny Mirror by Mario Klingemann is another work that I found really interesting, especially with the mirror reflecting not only your own image back but the faces of all previous audience viewers too. https://underdestruction.com/2020/08/29/uncanny-mirror/ |
| jhpa2017 |
I really enjoyed Live Brush to GauGAN thing https://twitter.com/fabinrasheed/status/1191255610479669248 As someone who does not consider himself an incredible artist, it is nice to see that there is an option for making a subpar painting into something a bit more realistic. It is astounding to me that something this advanced exists in the world of painting and art AI] I am also intrigued by the infinite drum machine (https://experiments.withgoogle.com/drum-machine) and the ability of computers to characterize and organize sounds in a way that we have no way of doing. |
| devon b (they/them) |
https://www.gwern.net/GPT-3 I really enjoyed this project attempting to create creative fiction through a GPT-3 model. Particularly enjoyed reading the failed attempts by the model to explain different puns. Along this line, if anyone hasn't played AI Dungeon I highly recommend it! It's a traditional text-based dungeon explorer game that generates the story as you play it. |
| HERE |
|
| Sophia Sun |
https://github.com/ml4a/ml4a/blob/master/examples/models/neural_style.ipynb I'm really interested in this project where the the creators try to recognize the patterns and styles in artworks and try to recreate artworks using a different style. Reminds me of art historians who might be able to discern which era the artwork belongs to just by analyzing the artwork styles and now AI is doing similar things! |
| simonalexander98 |
I really liked "Everyone Dance Now" https://www.youtube.com/watch?v=PCBTZh41Ris as it was a fun implementation of machine learning that worked quite well. It is amazing that machines can transform someone with minimal dancing abilities to look like a trained professional dancer! |
| chuck |
https://sougwen.com/project/drawingoperations-memory I really enjoyed this piece, found it incredible that the artist and robotic arm were able to work synchronously |
| Rafael |
https://www.aiweirdness.com/the-neural-network-has-weird-ideas-16-03-05/ Although some of the other projects were more technically ambitious, this one really made me laugh at the AI-generated recipes. |
| gabrielle-ohlson |
big fan of ‘learning to see’ from the first link—I think it is both very funny how much it calls humans out on their biases (in a weird and exciting way) and also a concept that I haven’t encountered before. well, I guess we’ve also seen people trying to humanize AI but I think this one has a quite adequate concept/explanation? http://www.memo.tv/works/learning-to-see/ |
| Sam |
https://www.youtube.com/watch?v=PCBTZh41Ris "Everybody Dance Now" was also my favorite because of how clean it looked. It seemed to do a really good job mapping the movements of the source material onto the person dancing. With a couple of exceptions that usually involved head turns, it was probably the cleanest application that I have seen of this technology so far. It seems like it could definitely be used in music videos where the artist isn't necessarily a professional dancer. |
| Pipi Gao |
My favorite ones are the AI brushes https://nurecas.com/ai-brushes especially the "character action brush". Transforming stick figures directly into characters is a lifesaver for those who can't draw (like me). The final part "9.Data Visualization" is also really interesting, since the output photo can be changed simply by manipulating data. |
| Jacob |
I really enjoyed checking out vibertthio's melodic ML projects https://vibertthio.com/sornting/ Very interesting to play around with the melody mixer and sornting minigame. |
| emersondorn |
https://www.youtube.com/watch?v=wvsE8jm1GzE |
| annaresek |
http://www.memo.tv/works/learning-to-see/ https://app.inferkit.com/demo https://sougwen.com/project/drawingoperations-memory |