Intel Urges AI Developers: ‘Go Big’

Intel® AI Academy DevJam Unleashes the Potential of AI

An image

At the recent AI DevJam event in New York, Intel challenged software developers, data scientists, and students to aim high as they take a unified approach to AI development on Intel® architecture … and presented them with the tools and resources to do it. This DevJam training featured a “first”: an event-within-the-event designed to engage and inspire AI developers, data scientists and grad students – the next generation of innovative Artificial Intelligence talent.

An imageAs a “Day 0” event preceding the New York O’Reilly AI conference, the AI DevJam attracted more than 600 registrants to take part in tech demos, poster sessions, fireside chats and other knowledge-building engagements. The objective was to introduce the array of Intel tools, frameworks, resources, and subject matter expertise to sharpen developers’ machine learning skills to deliver insights more quickly, solve analytics challenges, and provide a unified approach to shape the future of AI.

AI Students take part in Student Ambassador Forum

The three-hour Student Ambassador Forum event during DevJam attracted a full house of attendees from more than seven universities represented, including Rutgers, NYC, New Jersey Institute of Technology, Columbia, PACE, Stevens Institute of An imageTechnology, and Roger Williams University. The forum featured some of Intel’s newest Student Ambassadors as well as Intel experts and engineers who discussed their research, projects, and Intel's latest efforts in the AI space.

The Forum generated a successful turnout and positive feedback from student ambassadors and attendees alike. As a result, Intel may host similar forums in conjunction with AI DevJam trainings throughout the year; the next one is scheduled for CVPR on July 21 in Hawaii.

Three Intel Student Ambassadors shared their AI applications, all utilizing Intel’s AI optimized frameworks or tools, for:

  • Functional Connectivity of Epileptic Brains – Panuwat Janwattanapong of Florida International University used EEG technology to monitor electrical activity produced by neurons in the brain, leveraging Intel distribution for Python to derive performance increases.
  • Sensor fingerprint: mobile identity – Srivignessh Pacham Sri Srinivasan of Arizona State University demonstrated how, by associating a fingerprint to a device, mobile applications can prevent access to known malicious mobile users. And, it provides websites a cookie-less way of tracking users without triggering permissions warnings.
  • Facial recognition – Pallab Paul and Tejeswar Tadi, both of Rutgers, demonstrated a mobile app that uses computer vision to acquire date about facial structures using TensorFlow, Keras, Open CV, Java and Python on Intel® Xeon Phi™ cluster. Watch a video interview with Pallab Paul.

Demos, Fireside chats and more

DevJam attendees were treated to more than a half-dozen AI demonstrations, such as the Vehicle Rear Vision demo from Intel Innovator Peter Ma, which uses an Intel® RealSense™ camera to gauge the distance from obstacles. Later in the DevJam eventAn image, Scott Apeland, Director of the Intel® Developer Programs moderated two Fireside Chats. The first Chat featured three Student Ambassador presenters sharing “Academic Insights on AI.” The second Chat titled “The Future of Intel and AI” included pannelist Julie Choi, AI Product Group’s Head of Marketing Julie Choi and ElMoustapha Ould-ahmed-vall, Intel Senior Principal Engineer.

In addition to the Forum and Fireside Chat contributions, all three Student Ambassadors participated in DevJam demo stations and hosted well-attended poster chats inspiring interaction from the developer and academic audiences throughout the evening.

All in all, Intel AI DevJam inspired AI developers to “GoBig” as they unleash the full potential of AI on Intel architecture.

Stay updated with the next event and training for AI, machine learning and deep learning here .

For more complete information about compiler optimizations, see our Optimization Notice.