Timelapse is a global, zoomable video that lets you see how the Earth has changed over the past 32 years. It is made from 33 cloud-free annual mosaics, one for each year from 1984 to 2016, which are made interactively explorable by Carnegie Mellon University CREATE Lab’s Time Machine library, a technology for creating and viewing zoomable and pannable timelapses over space and time.
Source : Timelapse – Google Earth Engine
To train Google’s artificial Q&A brain, Orr and company also use old news stories, where machines start to see how headlines serve as short summaries of the longer articles that follow. But for now, the company still needs its team of PhD linguists. They not only demonstrate sentence compression, but actually label parts of speech in ways that help neural nets understand how human language works. Spanning about 100 PhD linguists across the globe, the Pygmalion team produces what Orr calls “the gold data,” w
Source : Google’s Hand-fed AI Now Gives Answers, Not Just Search Results | WIRED
« The AI vastly outperformed a professional lip-reader who attempted to decipher 200 randomly selected clips from the data set. The professional annotated just 12.4 per cent of words without any error. But the AI annotated 46.8 per cent of all words in the March to September data set.
The AI system was trained using some 5000 hours from six different TV programmes, including Newsnight, BBC Breakfast and Question Time. In total, the videos contained 118,000 sentences ».
Source : Google’s DeepMind AI can lip-read TV shows better than a pro | New Scientist
« Just in time for the Black Friday swarms, we’re adding a real-time look at how crowded a place is right now, to help you decide where and when to go. Whether you’re rushing to pick up a last minute gift or seeking a lively bar for some festive spirit, check Popular Times for a sneak preview of what to expect when you arrive ».
Source : Know before you go, with Google
At the start, we pioneered large-scale statistical machine translation, which uses statistical models to translate text. Today, we’re introducing the next step in making Google Translate even better: Neural Machine Translation. […]
At a high level, the Neural system translates whole sentences at a time, rather than just piece by piece. It uses this broader context to help it figure out the most relevant translation, which it then rearranges and adjusts to be more like a human speaking with proper grammar.
Source : Found in translation: More accurate, fluent sentences in Google Translate
Le débat est vif quant à la responsabilité de Google, Facebook et Twitter dans le résultat de l’élection de Donald Trump à la présidence des États-Unis. Ce débat est loin d’être terminé, tant il soulève des enjeux d’une rare complexité. La dernière en date porte sur la présence d’une information factuellement fausse dans les premiers résultats proposés par Google à la requête « final election numbers ».
C’est manifestement le cas depuis de nombreuses configurations (navigateur, IP,…), dont la mienne (cf. supra).
Beaucoup d’analyses, trop nombreuses, ne perçoivent pas que cette critique ne remet pas en cause ce résultat, mais le fonctionnement global de Google. Plus précisément, cette critique consiste à déplacer la pertinence des résultats de Google de l’autorité à la vérité. Or, la vérité est un problème autrement plus complexe que l’autorité, et Google, comme Facebook, ne vont probablement pas accepter cette responsabilité.
À présent, la démarcation entre média et plateforme se fait de plus en plus criante, et révèle les enjeux sous-jacents d’une telle distinction, dont la dimension politique ne pourra pas être évacuée indéfiniment.
Fil Menczer, a professor of informatics and computing at Indiana University who has studied the spread of misinformation on social media, said Google’s move with AdSense was a positive step. »One of the incentives for a good portion of fake news is money, » he said. « This could cut the income that creates the incentive to create the fake news sites. »However, he cautioned that detecting fake news sites was not easy. « What if it is a site with some real information and some fake news? It requires specialized knowledge and having humans (do it) doesn’t scale, » he said.
Source : Google, Facebook move to restrict ads on fake news sites | Reuters
With “RAISR: Rapid and Accurate Image Super-Resolution”, we introduce a technique that incorporates machine learning in order to produce high-quality versions of low-resolution images. RAISR produces results that are comparable to or better than the currently available super-resolution methods, and does so roughly 10 to 100 times faster, allowing it to be run on a typical mobile device in real-time. Furthermore, our technique is able to avoid recreating the aliasing artifacts that may exist in the lower res
Source : Research Blog: Enhance! RAISR Sharp Images with Machine Learning
Facebook could reach people not even thinking about a job, yet could be convinced to apply for the chance at a higher salary. That’s the same reason Facebook has been able to build giant brand and performance advertising businesses despite Google’s traditional dominance. You might search on Google when you know what you want to buy, or on LinkedIn if you know you want a job, and their ads can help with demand fulfillment. Facebook offers demand generation, drumming up interest people didn’t know they had.
Source : Facebook threatens LinkedIn with job opening features | TechCrunch
© 2025 no-Flux
Theme by Anders Noren — Up ↑
