In the News
China's rise in Artificial Intelligence
The country’s universities and tech giants are starting to surpass American ones when it comes to researching and implementing AI.
China is also taking the lead in the A.I. arms race with the U.S.
Amazon’s Alexa isn’t the future of AI
"It’s a glorified radio clock and, otherwise, stupid"
Google Acquisitions in 2017
Google will be continuing to look for Cloud and AI Deals in 2017. This is an interview with a Google Corporate Development VP on they way they choose, target and valuate projects in AI.
Pinterest introduces Lens, a Shazam for objects in the real world
Pinterest has been doing some impressive work in the Computer Vision field. Lens is its latest addition. More details on Visual Discovery at Pinterest in this paper.
Pinterest also recently hired a Google Image Search veteran. To be followed.
Also in the news...
- Ford invests a whopping $1B in self driving startup Argo AI. More
- Elon Musk thins (again) that Humans need to merge with machines. More
- Axon acquired Dextro for their video analysis technology. More
- Silicon Valley Hedge Fund takes on Wall Street with AI trader. More
Learning
Attacking machine learning with adversarial examples
"Adversarial examples are inputs to machine learning models that an attacker has intentionally designed to cause the model to make a mistake; they're like optical illusions for machines. In this post we'll show how adversarial examples work across different mediums, and will discuss why securing systems against them can be difficult."
Oxford Deep NLP 2017 course
Software tools & code
Announcing TensorFlow 1.0
Google has finally released the production-ready version of Tensorflow. Many changes, including some interesting speed improvements, and various improvements to the flexibility and consistency of the framework.
A few days earlier, Google also announced Tensorflow Fold with support for dynamic computation graphs.
YouTube-BoundingBoxes Dataset
New dataset with 5 million bounding boxes spanning 23 object categories, densely labeling segments from 210,000 YouTube videos.
Fully Convolutional Networks for Image Segmentation
A post showing how to perform Image Segmentation using Fully Convolutional Networks that were trained on PASCAL VOC using our framework.
An even easier introduction to CUDA
Creating massively parallel code for GPUs in C++. Useful beyond A.I.
Hardware
Next-generation GPU access on the web
Apple proposes new web standards to let web pages use the smartphone's GPU for Machine Learning and Graphics.
Some thoughts
'AI Brain scans'
Bristol-based Graphcore has used its AI processing units and software to create maps of what happens during a machine learning process.
About
This newsletter is a collection of AI news and resources curated by @dlissmyr. If you find it worthwhile, please forward to your friends and colleagues, or share on your favorite network!
Share on Twitter · Share on Linkedin · Share on Google+
Suggestions or comments are more than welcome, just reply to this email.
Thanks!