In collaboration with Google, we're releasing Activation Atlases: a new technique for visualizing what interactions between neurons can represent.

?Blog: https://t.co/nLfd0vpED4
?Paper: https://t.co/qh2gozddHE
?Code: https://t.co/VwLN5j3uZo
?️Demo: https://t.co/9d1PmPqt6H pic.twitter.com/7c6Q6GvNLq

— OpenAI (@OpenAI) March 6, 2019

Edge TPUs are getting closer to general release!! https://t.co/PCFrze5x5v (as of a month ago restrictions to send to Australia ; but I should recheck)

— mat kelcey (@mat_kelcey) March 5, 2019

For anyone that missed the "stealth" launch of #GoogleCoral yesterday, ahead of the the start of the TensorFlow Dev Summit tomorrow. Announced last year @Google's #EdgeTPU hardware is now on sale, and I have hardware on my desk! #machinelearning #edge #iot https://t.co/uvc3TzvogC pic.twitter.com/EyFVBk3vuX

— Alasdair Allan (@aallan) March 5, 2019

Say “hello” to Coral, @Google’s custom ASIC for machine learning on the edge: https://t.co/5neq0lvmde pic.twitter.com/ya8SNEiYPE

— Hackster Pro (@HacksterPro) March 4, 2019

Edge-y?? "Say “Hello” to Google Coral" – Hackster Blog - https://t.co/fB2FVc0Jou #ML #edge

— highendcompute (@highendcompute) March 6, 2019

Say hello to Google Coral (Ege TPU) https://t.co/uw33MBK7Eq pic.twitter.com/0d3b08CsUt

— Jun Rekimoto : 暦本純一 (@rkmt) March 6, 2019

soon everyone will have a machine learning hardware accelerator or TPU just like how we all consider GPUs a part of a normal pc that runs all the latest good stuff.

"make sure you update to the latest AI drivers to play this new game"https://t.co/oPdmZarMyh

— Christer McFunkypants Kaitila (@McFunkypants) March 6, 2019

…the official #GoogleCoral announcement is now up on the @GoogleDevs blog, https://t.co/l2RQFiPxir. Full details of @Google's new #EdgeTPU hardware in my piece for @HacksterIO. #MachineLearning #Edge #IoT #TFDevSummit https://t.co/uvc3TzvogC

— Alasdair Allan (@aallan) March 6, 2019

The original #Movidius Neural Compute Stick (left), the newer rebranded #Intel Neural Compute Stick 2 (middle), and the #Google Coral USB Accelerator (right) released yesterday, https://t.co/uvc3TzvogC. You know where this is heading, don't you..? ? #machinelearning #benchmarks pic.twitter.com/oqCedHJzkB

— Alasdair Allan (@aallan) March 5, 2019

"Say 'Hello' to Google Coral," by me for @HacksterIO. Announced last year, @Google's #EdgeTPU hardware silently launched into Beta earlier today under the name #GoogleCoral. Both the USB accelerator and the dev board are now shipping! #machinelearning #iot https://t.co/uvc3TzvogC

— Alasdair Allan (@aallan) March 4, 2019

Was already excited for the #TFDevSummit and now this come out: Say ‘Hello’ to Google Coral by @aallan https://t.co/Qj3almuuix #DeepLearning is coming for everyone!!!

— mimmo iaia (@mimmoiaia) March 5, 2019

Google is making it easier for AI developers to keep users’ data private https://t.co/PFo0OA4QAb pic.twitter.com/tq7Q6j373f

— The Verge (@verge) March 6, 2019

Google releases TensorFlow Privacy note that there are tradeoffs here and it won't fit every application: "By masking outliers, it can sometimes remove relevant or interesting data, especially in varied datasets, like those involving language" https://t.co/kLL4aUOCsf

— Peter Skomoroch (@peteskomoroch) March 6, 2019

Google announces TensorFlow Privacy, an open source module for the machine learning framework that lets developers safeguard data with differential privacy (James Vincent/The Verge) https://t.co/skRwn4whWw

— Chris Heilmann (@codepo8) March 6, 2019

Introducing Coral: Our platform for development with local AI https://t.co/9mxgOy5GsE

— Chris Heilmann (@codepo8) March 6, 2019

Introducing Coral: Our platform for development with local AI https://t.co/fABmrfCXeW #News #Google platform #tools

— The Guru (@softwarereview7) March 6, 2019

ちょっとだけ楽しみ / Google Developers Blog: Introducing Coral: Our platform for development with local AI https://t.co/xALv0ggQCx via @google

— ??????? シアラナ (@cia_rana) March 6, 2019

TensorFlow Privacy, using techniques based on differential privacy, will make it easier for developers to train ML models with privacy and better protect users' data in their AI development. #GoogleAI https://t.co/niG8Q6nyye

— Sundar Pichai (@sundarpichai) March 6, 2019

Our @GoogleAI team released a blog post from @schmilblick42 and @UlfarEr about @TensorFlow Privacy: the library is the result from a large collaboration and makes it easier to train TF models with strong privacy guarantees: https://t.co/7lO4F4drwC Excited to see it being used! https://t.co/U3g8GqdBwg

— Nicolas Papernot (@NicolasPapernot) March 6, 2019

Google unveils new tools to bolster AI hardware development https://t.co/gOexB8kdh5 pic.twitter.com/qejmAPfVMW

— The Verge (@verge) March 6, 2019

GOOGLE LAUNCHES AI PLATFORM THAT LOOKS REMARKABLY LIKE A RASPBERRY PIhttps://t.co/GNfpT4xqVk#RaspberryPi #Arduino #computers #Gadget #technology #Robots #Robotic #IoT #education #STEM #STEMeducation #TechnologyNews pic.twitter.com/7LFXVGkZHY

— vilrosusa (@vilrosusa) March 6, 2019

You connect the Coral through USB port
It contains a functional TPU board
It’s eighty bucks only
For inference, mostly
I’m buying one, life is too shorthttps://t.co/ICcxYQtNOR

— ML Limericks (@MLimericks) March 6, 2019

Google Launches AI Platform That Looks Remarkably Like A Raspberry Pi via @hackaday https://t.co/tPW6j9S2YM

— Denis Skinner (@SkinnerDenis) March 6, 2019

Google의 AI 선물세트들 https://t.co/9XKySlMUMf

— editoy (@editoy) March 6, 2019

Googleがラズパイぽいの出すんですね。日本からはまだ買えないみたいですが、アメリカだと$150。正式にはTensorFlow Dev Summitで発表のようですが、すでに1300のプリオーダーがwhttps://t.co/VXOIn56pUBhttps://t.co/2Vvhb9HFv7

— しょーご (@shogonishimoto) March 7, 2019

Between this device and Google Coral (https://t.co/UuqDbTMzuv), creating embedded ML experiences on the edge, has become highly accessible. Opens up many interesting use-cases. https://t.co/O4CWrBA32o

— samim (@samim) March 7, 2019

Really cool demonstration of on-device transfer learning on the #GoogleCoral Dev Board by @jtgans, https://t.co/uvc3TzvogC. #TFDevSummit #Edge #IoT #MachineLearning https://t.co/zK8HsPLSDI

— Alasdair Allan (@aallan) March 6, 2019

Google Coralから EDGE TPUを搭載した開発ボードと USBアクセラレータの発表。購入も可能 #google #ML https://t.co/kCceSgDslG

— Tatsuo Nagamatsu (@nagamatu) March 7, 2019

Say ‘Hello’ to Google Coral by @aallan https://t.co/EdDH8MhJpX

— Park JaeHo (@jrogue) March 6, 2019

Edge TPU, Google’s custom ASIC for machine learning has arrived ~> By @aallan https://t.co/KnNi1wZJAU

— Soraya Paniagua ✏️ Periodista (@sorayapa) March 6, 2019

Google is making it easier for AI developers to keep users’ data private https://t.co/PFo0OA4QAb pic.twitter.com/OvADzXDobQ

— The Verge (@verge) March 7, 2019

Google is making it easier for #AI #developers

to keep users’ #data #private https://t.co/RDZPVmuYxj #fintech #insurtech #ArtificialIntelligence #MachineLearning #DeepLearning #robotics #privacy @jjvincent #BigData @verge pic.twitter.com/YdqFEixMZd

— Spiros Margaris (@SpirosMargaris) March 7, 2019

Google サイト投稿記事(3/6)"Introducing Coral: Our platform for development with local AI" (https://t.co/IpWOeDUbT6)を読んで、"Coral Board"は"DL accelerator搭載 Raspberry Pi Zero W"のような印象を持った。AIカメラの試作には便利かも。

— Hiroyuki Kami (@Kamoike) March 7, 2019

Google Developers Blog: Introducing Coral: Our platform for development with local AI https://t.co/Z4wSTsOwlr @googleさんから

— ソースコード解析職人 (@Vengineer) March 7, 2019

Tensorflow dev summit is kicking off in 5 min, gonna be an exciting day!
Some stuff from my team:
TF differential privacy module https://t.co/Ws3fbZoyiy
Tensorflow Federated https://t.co/hB2M87sHPV

— Stefan Dierauf (@sdierauf) March 6, 2019

Google tool lets any AI app learn without taking all your data https://t.co/kemTr0LSsR via @CNET

— brent hoberman (@brenthoberman) March 6, 2019

Google tool lets any AI app learn without taking all your data https://t.co/C1WfQzrPyg

— Laura Hautala (@lhautala) March 6, 2019

Google's new TensorFlow Privacy library allows developers to train machine-learning models with differential privacy. https://t.co/OplfBp1zri

— SitePoint (@sitepointdotcom) March 7, 2019

Privacy and #machinelearning #deeplearning #technology #ai #art #artificialintelligence https://t.co/x2pUwP4rLY

— KM JOBS (@kmjobs) March 7, 2019

Google unveils new tools to bolster AI hardware development#ArtificialIntelligence #AI #MachineLearning #TechNews #technology #deeptechhttps://t.co/25Wca3Gv1b pic.twitter.com/5cMs9iDsA4

— DeepTechWire (@DeepTechWire) March 6, 2019

The next advancement in deep learning isn't a new algorithm, framework, or language. It is low-power edge computing devices that allow on-device inference at a low cost. #MachineLearning #DeepLearning #AI #computervision
https://t.co/k1qbU4l3YF

— Adam Mikulis ?️ (@adammikulis) March 6, 2019

Google Launches AI Platform That Looks Remarkably Like A Raspberry Pi via @hackaday https://t.co/OxgCWw1ulZ

— Michael Young (@Jaraxal) March 6, 2019

Google IoT AIプラットホームをアナウンス

AIシングルボードコンピュータとAI USBアクセラレータ

>Google announces Coral platform for building IoT hardware with on-device AI - 9to5Google https://t.co/ZihfCdJAzA

— T.Fki (@tfki) March 6, 2019

Google begins selling the $150 Coral Dev Board, a hardware kit for accelerated AI edge computing https://t.co/ntPbTsb1dz @VentureBeatさんから

— ソースコード解析職人 (@Vengineer) March 7, 2019