The trained skeleton has a cheerful gait: an entry to the “Learning to Run” competition, by modeling muscle excitations for hamstrings, biceps femoris, gluteus maximus, & others.
What jobs and tasks can AI actually automate? A new initiative from the University of Oxford is inviting you to answer this question – creating a new database of automatability predictions. You might recognize the researchers as Frey & Osborne, whose 2013 paper ignited the recent discussion on automation.
News & Articles
The city of Toronto has begun data-sharing with traffic app Waze in order to reduce congestion. Traffic congestion authorities and apps throughout are beginning to team up.
Papers & Tutorials
Computer Vision paper from Apple: 3D bounding boxes in 3D point clouds. Similar to our FusionNet paper, but with bounding boxes. Operates on the raw point cloud. Looking forward to more Machine Learning papers from Apple.
Imposing custom emotions and facial characteristics on faces using Generative Adversarial Networks. Can make input faces angry, contemptuous, disgusted, fearful, happy, sad, surprised, or even change gender.
Yann LeCun keynote at the first Cognitive and Computational Neuroscience conference at Columbia University on September 6, 2017.
Videos from Cognitive Computational Neuroscience (CCN) 2017
Talk with Peter Donnelly in his capacity with the Royal Society’s Machine Learning Working Group about the work they’ve done on the public’s views on AI and ML.