Travis Feldman

After my discussions with VP Pence and others, the federal government has agreed to withdraw federal officers from Portland. They have acted as an occupying force & brought violence. Starting tomorrow, all Customs and Border Protection & ICE officers will leave downtown Portland.

The Robotic Modeling Assistant (RoMA), an interactive fabrication system providing a hands-on and in-situ modeling experience with an augmented reality CAD editor and a robotic arm 3D printer. #tech #3dprinting #robotics #ar #engineering

"Pattern Recognition and Machine Learning" by @ChrisBishopMSFT is now available as a free download. Download your copy today for an introduction to the fields of pattern recognition & machine learning: aka.ms/prml #ML #Insights

Important paper from Google on large batch optimization. They do impressively careful experiments measuring # iterations needed to achieve target validation error at various batch sizes. The main "surprise" is the lack of surprises. [thread] arxiv.org/abs/1811.03600

A fun dataset to try new things? 🏫

Excited to share #Creatability, a new collection of experiments exploring ways to make creative tools more accessible, made w/ friends @AbilityNYU @JAZDeafMusical @openupmusic @ChanceyFleet @BerkeleyBlink & more. Built w/ @TensorFlow Play with them here → g.co/creatability

Introducing the MAESTRO dataset (172 hours of aligned MIDI and audio piano performances) and Wave2Midi2Wave: a factorized approach to modeling and generating piano performances. w/ @fjord41 Andriy Stasyuk @iansimon @huangcza @sedielem Read 👉 g.co/magenta/maestr… Listen 👇

UX vs UI

We are trying to get these Machine Learning cheatsheets accessible to more people around the world. If you want to help us with the translation into your own native language, you can contribute in this dedicated repo: github.com/shervinea/chea…

Remembering steel guitar pioneer Pete Drake, born on this day in 1932 in Augusta, Georgia.

Reinforcement Learning for Improving Agent Design: What happens when we let an agent learn a better body design together with learning its task? article: designrl.github.io pdf: arxiv.org/abs/1810.03779

Amazing find from the Getty Images archives: high resolution footage of someone playing a computer game in *1968*

Gamelan recorded in Peliatan, Bali, Indonesia in 1985.

New paper on regularizing adversarial learning with a variational discriminator bottleneck. Our method (VAIL) can imitate challenging skills from mocap data, and can even learn to run directly from a video clip: xbpeng.github.io/projects/VDB/i…

common ancestor for #stagbeetle and #dinosaurs? #BIGGAN

Had an idea to make my AR trajectory prediction more accurate! Now simulates until where it will land, updating prediction every frame

Photographer Lewis Hine (born #onthisday 1874) travelled the US from 1908 to 1911 documenting child labor. Among his 100s of photos is a set of composite images of Southern cotton mill workers, several workers overlaid on the same photographic plate buff.ly/2xPVVWf #OTD

I trained a pix2pix model with the Stanford dog image dataset. It seems to have learnt that a large U-shape line can be a tongue. github.com/keijiro/Pix2Pix

Cat-to-cat translation with pix2pix post fx. It's implemented as a custom effect for the post-processing stack. github.com/keijiro/Pix2Pix

Real time face segmentation in the browser ... Demo:zaidalyafeai.github.io/face-segmentat… Code :github.com/zaidalyafeai/z…