Sponsor
Attend the Future Labs AI Summit in NYC on October 30 – 31
Two days of technical trainings and talks from leading executives at Google, Vector Institute, NYU, Insight Data Science, AI4ALL, and more. Courses in machine learning, game theory, and the implementation of sophisticated techniques. Register today before prices go up on October 25!
In the News
How to Regulate Artificial Intelligence
Machine-vision drones monitor animals in the African savanna
Managing wild animals in remote areas requires accurate estimates of their numbers. Machine-vision drones can help.
Also in the news...
- Microsoft AI research has grown to 8K people. More
- Matroid raises $10M for easy-to-use video analysis tools. More
- Fast.ai, an online AI-course directory, decided to switch from Tensorflow / Caffe to Pytorch. Might have long-term implications on the usage of those different frameworks. More
Workplace
Transitioning from Academic Machine Learning to AI in Industry
Learning
Theory cracks open the Black Box of Deep Learning
A new idea called the “information bottleneck” is helping to explain the puzzling success of today’s artificial-intelligence algorithms — and might also explain how human brains learn.
Object detection: an overview in the age of Deep Learning
Great rundown on principles such as classification, localization, segmentation, along with their uses and their main implementations (Faster R-CNN, YOLO, SSD and so on).
3D Face Reconstruction with CNNs
Interesting research on large-pose 3D face reconstruction from a single image. To be compared to previous approaches that used an array of images to reconstruct the geometry.
Software tools & code
Unity Machine Learning Agents
Unity, a hugely popular video-game creation engine introduces learning agents to help drive development in Reinforcement Learning Algorithms.
Matterport3D Research Dataset
Dataset with 10,000 aligned and labeled panoramic views (RGB + depth per pixel). Will this be the new Imagenet?
Hardware
Intel presents a neuromorphic chip (research)
Loihi research test chip includes digital circuits that mimic the brain’s basic mechanisms (neuromorphic chip), enabling ML training and inference directly on chip. This should make machine learning faster and more efficient while requiring lower compute power. The chip is built with Intel's 14nm process technology and has 130,000 neurons and 130 million synapses.
Some thoughts
Why AI companies can’t be lean startups
Discussion with Matt Turck about the specificities of AI companies and how they compare to there SaaS predecessors.
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!