We invented technology to enhance the fan/performer connection. Vote for Team Ambience at EDC!
DJs and more traditional musicians require realtime audience feedback during performances. However, often we cannot see our audience—their movement, their facial expressions, etc.—during shows due to stage lighting. Therefore we cannot gauge their enthusiasm, and therefore cannot alter our performance to respond.
We at Team Ambience invented a means for measuring such audience enthusiasm using a video camera and artificial intelligence (AI). Our solution reports mood metrics and summarizes crowd movement directly to performers’ phones. (Many musicians, myself included, perform with an iPhone connected to our mic stands).
We developed our prototype from scratch while attending the Amazon- and Intel-sponsored “Hack Till Dawn – A Deep Learning Hackathon: Use Machine Learning to Revolutionize EDC Festivals” event last weekend. At this hacking competition, teams competed to win tickets to the Electric Daisy Carnival (EDC) by developing AI technologies designed to improve fan experiences at large music festivals. By strengthening performers’ ability to “read” the audience in realtime, our invention definitely enhances the fan experience by enabling performers to better craft their shows according to audience sentiment.
EDC will feature Team Ambience’s (our) solution, along with those of two additional winning teams, at three kiosks positioned throughout the festival this coming weekend. Fans will vote on their favorite idea—we certainly encourage you to vote for us!
Reading Audience Facial Expressions
The first video below shows us presenting different facial expressions to performers. The phone in the lower right reports our mood:
Zooming in on the phone application itself, note the change as the music changes:
Reading Audience Movement
Is the audience jumping, are they swaying? If so, how much? Consider the following video :
From this we measure audience movement intensity and report it:
The audience metrics described hold value not only to performers, but also to venue owners, booking agencies, and artist managers. We provide a business dashboard for such stakeholders:
Furthermore, as a composer/performer, I envision creating an AI that composes and improvises with me on-the-fly using audience feedback.
We used the Amazon machine learning stack, along with a AWS DeepLens camera:
We could not have achieved this without a complimentary AWS DeepLens camera (thanks Amazon!). This wireless video camera contains a dedicated machine learning chip, along with preloaded algorithms. Furthermore, you can add your own algorithms. The whole solution is tightly integrated with the rest of the AWS machine learning stack. Click on the image below to go to the DeepLens page. (Disclosure: I am an affiliate marketer for Amazon).
- I apologize that I do not know the source of this video.