A cool paper from Facebook AI (not from FAIR!) about detecting and reading text in images, at scale.
— Yann LeCun (@ylecun) July 24, 2018
This is very useful for detecting inappropriate content on Facebook.
The system uses R-CNN/Detectron for... https://t.co/wfchAeDgo3
The difference between the ground truth samples and their output is (to my ear) undetectable. I am either misinterpreting their example clips or this is remarkable. https://t.co/KLznHi7qzM
— iandanforth (@iandanforth) July 24, 2018
Samples from CPPN-GAN trained on Chinese handwriting dataset https://t.co/U1wc8qXLME pic.twitter.com/1rXwUttH5z
— hardmaru (@hardmaru) July 24, 2018
Our paper on probabilistic programming for large-scale scientific simulators is out: https://t.co/46ZNuB0fe5
— Kyle Cranmer (@KyleCranmer) July 23, 2018
pyprob is like pyro in @PyTorch but w/ a protocol for simulators in other languages https://t.co/TtIjD4o96nhttps://t.co/0YkXw00VCo
Awesome work @atilimgunes @tuananhle7! pic.twitter.com/zVaZDmOdvj
Very nice paper on Cartesian Genetic Programming applied to Atari games. https://t.co/f08gWRpP9S -- State-of-the-art on a number of games, and very competitive on many more.
— François Chollet (@fchollet) July 22, 2018
New in @epj_ds
— Alessandro Vespignani (@alexvespi) July 21, 2018
Mining large-scale human mobility data for long-term crime prediction
“large human mobility data to craft an extensive set of features for crime prediction, as informed by theories in criminology and urban studies. “https://t.co/GQKfHO7eZD pic.twitter.com/nK02uY3qkG
Played around with @dpkingma & @prafdhar interactive 'Glow' reversible generative model and was able to knock a few years off of Solid Snake and dye his hair.
— Jack Clark (@jackclarkSF) July 21, 2018
Blog: https://t.co/HjKlZ8okkZ
Code: https://t.co/U80xSUOUrC pic.twitter.com/LxB6l9TSCT
Code for the Glow demo is now on github! Use the pre-trained model to create manipulation vectors from your own labelled dataset (eg: changing hair color to red). You can also run the demo locally, go make your own smiling videos!https://t.co/kLTURsLX0d pic.twitter.com/05T3gzs9gg
— Prafulla Dhariwal (@prafdhar) July 21, 2018
Deep RL, from images, conditioned on user-specified goals, trains in a few hours on real robots:https://t.co/317cmDyYph
— Sergey Levine (@svlevine) July 20, 2018
Key idea: let the robot "practice" goals that it proposes to itself to learn more about the world.
w/ @Vitchyr Ashvin Nair, M. Dalal, S. Bahl, S. Lin
Modular networks, modules transform representations to get the answer, by chaining together a sequence of analogies: https://t.co/aJ9C8nuzK4
— Sergey Levine (@svlevine) July 20, 2018
Model "thinks out loud" with intermediate representations like "8 usplay evensay inusmay nine"
w/ @mmmbchang A.Gupta and T.Griffiths
The final version of our paper is now online!
— Battista Biggio (@biggiobattista) July 20, 2018
Wild Patterns: Ten Years after the Rise of Adversarial Machine Learning
To appear on the 50th Anniversary Edition of Pattern Recognition: https://t.co/sOrRBeFf78
Have a look at our updated review and timeline of attacks against ML! pic.twitter.com/iWriFelxaU
Check out this ICML tutorial for a whirlwind tour of current Imitation Learning research: https://t.co/5BWgKuh2vX
— Denny Britz (@dennybritz) July 20, 2018
