Qualcomm shows off AI-equipped car at CES 2019 and more from TWiML & AI

1024 768 This Week in Machine Learning & AI

Bits & Bytes

  • Google introduces Feast: an open source feature store for ML. GO-JEK and Google announced the release of Feast that allows teams to manage, store, and discover features to use for ML projects.
  • Amazon CEO, Jeff Bezos, is launching a new conference dedicated to AI. The new AI specific conference, re:MARS, will be held in Las Vegas between June 4th and 7th this year. Should be an interesting event.
  • Mayo Clinic research uses AI for early detection of silent heart disease. Mayo Clinic study finds that applying AI to electrocardiogram (EKG) test results offers a simple, affordable early indicator of asymptomatic left ventricular dysfunction, a precursor to heart failure.
  • Microsoft announces ML.NET 0.9. Microsoft’s open-source and cross-platform ML framework, ML.NET, was updated to version 0.9. New and updated features focus on expanded model interpretability capabilities, GPU support for ONNX models, new Visual Studio project templates in preview, and more.
  • Intel and Alibaba team up on new AI-powered 3D athlete tracking technology. At CES 2019, Intel and Alibaba announced the new collaboration to develop AI-powered 3D athlete tracking technology to be deployed at the 2020 Olympic Games.
  • Baidu unveils open source edge computing platform and AI boards. OpenEdge, an open source computing platform enables developers to build edge applications with more flexibility. The company also announced new AI hardware development platforms BIE-AI-Box with Intel for in-car video analysis, and BIE-AI-Board, co-developed with NXP, for object classification.
  • Qualcomm shows off an AI-equipped car cockpit at CES 2019. At CES, Qualcomm introduced the third generation of its Snapdragon Automotive Cockpit Platforms. The upgraded version covers various aspects of the in-car experience from voice-activated interfaces to traditional navigation systems. Their keynote featured a nice demo of “pedestrian intent prediction” based on various computer vision techniques including object detection and pose estimation.

 

Dollars & Sense

To receive the Bits & Bytes to your inbox, subscribe to our Newsletter.

Leave a Reply

Your email address will not be published.