Next week join experts at Intel Vision for sessions covering all things AI, including how it’s being used to solve complex manufacturing processes, improve remote patient monitoring, and help enterprises automate and digitize non-core functions so they can focus on delivering products and services.
Efficiency, Extensibility, and Cognition: Charting the Frontiers of Deep Learning
Even as DL continues to drive technological progress, several fundamental limitations must be overcome so AI in general can realize its full potential. Gadi Singer explores three key areas where Intel Labs’ Cognitive Computing Research is working to extend innovation beyond the limits of contemporary DL.
How MLOps Is Helping Overcome Machine Learning’s Biggest Challenges
Many enterprise ML struggles boil down to two things: processes and infrastructure. MLOps—which applies the principles of DevOps to ML—reduces friction between development and engineering teams, leading to increased productivity, faster model training, and better compute utilization.
Intel Wins DARPA RACER-Sim Program
DARPA has awarded Intel, Barcelona’s Computer Vision Center, and UT Austin a chance to develop advanced simulation solutions for off-road autonomous ground vehicles. This RACER-Sim program aims to significantly reduce development costs and bridge the gap between simulations and the real world.
Computer Vision Solution Keeps Private Data Out of the Picture
Pattern recognition systems in CV and audio processing are usually based on explicit models with multilayer hierarchical feature extractors. Learn how Carnegie Mellon University and Intel Labs developed a one-stage implicit deep model that can scale to realistic visual tasks using less memory.