Episode 3: Life is like a game

Video games have become a favourite tool for AI researchers to test the abilities of their systems. Why?Read More

Related Google News:

SRE at Google: Our complete list of CRE life lessons April 27, 2021
What no-code automation looks like with AppSheet April 20, 2021
Bring iconic Japanese characters to life with AR in Search April 20, 2021
Churn prediction for game developers using Google Analytics 4 (GA4) and BigQuery ML April 14, 2021
Seeking aspiring designers looking to Change The Game April 7, 2021
Picture this: 10 whiteboard sketch videos that bring Google Cloud to life March 24, 2021
High Performance Game Audio with Oboe March 23, 2021
Leveraging Machine Learning for Game Development March 19, 2021

August 19, 2019

AI / Google

Google AIFollow

Google AI@GoogleAI·

6 May

Congratulations to the Googlers recognized as new @sig_chi Academy Members at the #chi2021 Awards Ceremony - Sunny Consolvo, Merrie Morris (@merrierm), Fernanda Viégas (@viegasf) and Allison Woodruff. Learn more at https://sigchi.org/awards/sigchi-award-recipients/2020-sigchi-awards/ and https://sigchi.org/awards/sigchi-award-recipients/2021-sigchi-awards/

Reply on Twitter 1390452781148901377 Retweet on Twitter 139045278114890137719 Like on Twitter 1390452781148901377142 Twitter 1390452781148901377

Google AI@GoogleAI·

6 May

Datasets that pair images with captions are widely used to build #DeepLearning captioning models, but these datasets’ missing cross-modal associations limit both research and practical applications. Learn how Crisscrossed Captions addresses the gap at https://goo.gle/3h2cRku

Reply on Twitter 1390413522341875714 Retweet on Twitter 1390413522341875714105 Like on Twitter 1390413522341875714461 Twitter 1390413522341875714

Google AI@GoogleAI·

5 May

Existing text-editing models are either fast or flexible, but not both. Learn about FELIX, a fast and flexible text-editing model that achieves state-of-the-art performance with a 90x speed-up compared to seq2seq approaches. Read more and grab the code → https://goo.gle/33lrCXo

Reply on Twitter 1390046279468830721 Retweet on Twitter 1390046279468830721116 Like on Twitter 1390046279468830721418 Twitter 1390046279468830721

Google AI@GoogleAI·

4 May

Today we present a systematic study comparing wide and deep #NeuralNetworks through the lens of their hidden representations and outputs. Learn how similarities between model layers can inform researchers about model performance and behavior at https://goo.gle/3xOlm8A

Reply on Twitter 1389680797985161218 Retweet on Twitter 1389680797985161218286 Like on Twitter 13896807979851612181210 Twitter 1389680797985161218

Google AI@GoogleAI·

3 May

Head over to the Google AI blog to see the research we are presenting at #ICLR2021, which kicked off this week. If you’re registered at this @iclr_conf, visit the online Google booth to learn even more about our research being presented! #DeepLearning https://goo.gle/2QPJWp3

Reply on Twitter 1389323269136961536 Retweet on Twitter 138932326913696153645 Like on Twitter 1389323269136961536200 Twitter 1389323269136961536

Google AI@GoogleAI·

29 Apr

Today we describe RecSim NG, a next generation simulator for recommender systems research and development that extends RecSim to address the gap between simulation and real-world applications. Learn more and grab the code at https://goo.gle/3e4D3sN

Reply on Twitter 1387883390570012674 Retweet on Twitter 138788339057001267476 Like on Twitter 1387883390570012674312 Twitter 1387883390570012674

Google AI@GoogleAI·

28 Apr

Introducing a model-based #RL approach for robot navigation, called hierarchical predictive planning (HPP), that enables agents to align their goals on the fly in order to solve the decentralized rendezvous task. Learn more at https://goo.gle/3sZSyqb

Reply on Twitter 1387459211496501253 Retweet on Twitter 1387459211496501253202 Like on Twitter 1387459211496501253861 Twitter 1387459211496501253

Google AI@GoogleAI·

27 Apr

Reconstructing the 3D world from 2D video requires monocular depth estimation and video panoptic segmentation, but they are typically considered separately. Today we present ViP-DeepLab, which does both at once while achieving state-of-the-art performance→https://goo.gle/2S9BVM0

Reply on Twitter 1387141759164194820 Retweet on Twitter 1387141759164194820467 Like on Twitter 13871417591641948202014 Twitter 1387141759164194820

Google AI@GoogleAI·

23 Apr

While burst photography enables the capture of high-quality images on mobile, it can result in noisy shadows in scenes with high dynamic range (HDR). Learn how HDR+ with Bracketing merges multiple frames with different exposures to improve image quality → https://goo.gle/3aBHp8B

Reply on Twitter 1385652787321602052 Retweet on Twitter 138565278732160205273 Like on Twitter 1385652787321602052370 Twitter 1385652787321602052

Google AI@GoogleAI·

22 Apr

Today we present a new approach for automated discovery of generalizable #ReinforcementLearning algorithms that evolves a population of loss functions — represented by computational graphs — over a set of simple training environments. Learn more at https://goo.gle/3dFzDML

Reply on Twitter 1385324565589491712 Retweet on Twitter 1385324565589491712302 Like on Twitter 13853245655894917121165 Twitter 1385324565589491712