
A showcase with creative machine learning artworks and projects. One new feature every day. Curated by @emilwallner Submit: https://teapawt.org/t.co/8Wd9NUyFE4
Paris, France
Joined October 2013
- Tweets37
- Following152
- Followers3K
- Likes1.3K
New feature on ML x Art: @advadnoun→ bit.ly/39Jwybx
Another fun exploration with MinMax trained on a few images on CelebA. Love the shapes and curves! 𑗊
NEW - @Guillaume_Slize's The Devogram 🤖bit.ly/2Y5UUGP
Furby has become *SENTIENT* 🥰
A mobile broadcasting robot using MobileNet V2 trained on COCO, OCR, and text to speech to amplify messages in public spaces.
NEW: (Non-)Human by @ygantigravity ☞ bit.ly/3rYrM1h
A GAN trained on the artists sleeping patterns and then interpolated by a custom sheet folding machine.
OMG!! ❤️
LATEST: Ouroboromorphism by @wavefunk_ → bit.ly/2ZoQxr2
Audio-reactive latent interpolation with StyleGAN2 created by modulating latent and noise inputs with musical features.
These textures are rad! Loving the latent music space! 🖤
NEW: Latent Flowers GANden by Zihou Ng 👉 bit.ly/2Zt6Ov3
It's fun to see @observablehq being used for ML explorations and benefiting from WebGL!!
We need more beautiful, hackable, tutorials with GUIs!
JUST IN: SYN FEATHER SWEATER by @synflux_io and @hatroid → bit.ly/3jW5Vor
Loving these sweaters with synthetic GAN birds converted to double Jacquard knit format.
I bet we'll see a few fellow nerds sporting these at the next @NeurIPSConf and @avant_galerie events. 🔝
NEW - @Guillaume_Slize's Algorithmic readings of Bertillon’s portrait parlé ☞ bit.ly/3qMMaBO
A character generator model is trained on a technical database to identify criminals from 1907. It's contrasted with a modern webcam linked to a facial recognition algorithm 🤯
JUST IN: 🪲 by @cunicode👉 bit.ly/3jpnOf8
A StyleGAN interpolation of beetle illustrations. The clean backgrounds and detailed drawings are ideal ingredients for a GAN.
These are LOVELY!
NEW: A Piano Choir by @yotammann ☞bit.ly/3q08vMr
There is something transcendent about listening to a singing piano, the disconnection between what you see and hear is wonderful.
youtube.com/watch?v=pnZQHd0I…
NEW: @sam_lavigne's collage of Mark Zuckerberg's haircuts using semantic segmentation → bit.ly/3cPkOXX
Looking for a fresh spring cut or a new wallpaper, look no further!!
Oh, Sam also open-sourced the dataset: lav.io/zuckerhair.zip 🏆
JUST IN: @Lialialiacole's flower fleshdigressions👉bit.ly/3cOYIF6bit.ly/3cOYIF6
There is something special with music-based GAN interpolations! Here, using @nathanchancello's excellent cello piece.
The blooming flowers are dancing and the palette is sublime!
teapawt.org/Lialialiacole/status/1…
Sound on! Enjoy my flower fleshdigressions synced up to some lovely cello music by @nathanchancello. 🥰🎶 Having lots of fun with TouchDesigner! #generative #StyleGAN2 #stylegan #generativeart
ML x Rock 'n Roll 🤘 → bit.ly/3aG8FC4
It's ace to see ML get more traction in music! Here, using a WaveNet model for guitar AMPs (audio distortion circuits).
AMP: @Guitar_ML1
Article: @ZakJost
Paper: mdpi.com/2076-3417/10/3/766/…
NEW📽: by @Triangle3Edge ☞ bit.ly/2LxkFgO
It's amazing what a few no-code ML tools can do to a video - cinematic!!
A pandora's box for indie moviemakers.
Source: bit.ly/36OKEaI
@deepart_io and portraitai for the style and EbSynth (@scrtwpns) for the video.
New artwork on ML x Art: @unityofmulti → bit.ly/39En29r
Stong pose made in @Artbreeder, and excellent glitch/cyberpunk effect with AICAN 🤯
NEW - Loss Landscape by @ideami 👉 bit.ly/2Y5NMu2
These are beautiful. AND insightful.
Javier Ideami has created a tool to visualize the loss function in ResNet models. 👌
NEW - Strange Attractions by @lenkahamosova & @pavolrusnak → bit.ly/2Mm8e7G
The glitchiness of this model is fabulous!
AND the COLORS!
NEW /// @Guillaume_Slize's I will remember → bit.ly/39V7DBX
I LOVE how this came together.
The anthropomorphism, clash, and lush setting!
LATEST: A StyleGAN portrait by @minneatairu → bit.ly/3cx44EY
Beautiful.
NEW - Octopus and Adderall by @bonkerfield👁🐙👁bit.ly/3ashtey
Generate a video from the artist's 3500-word story using CLIP-based Story2Hallucination.
This is a glimpse into the feature - WOW!
JUST IN - Near Miss by @annespalter → bit.ly/3o91W8j
An image Style Transferred with pictures of three airplanes using @NightcafeStudio and the spaceman and asteroid are animated with @Plotaverse.
The shapes and smoothness in the animations are mesmerizing!



