Computer Vision with OpenVINO
Without writing one line of code and thanks to the OpenVINO toolkit, any Intuiface experience can now be notified about the age range, gender, head pose, dwell time, and emotional state of each person in front of a screen - all at no cost.
Intuiface and Intel cohosted a live deep dive about our offering of face detection with help of OpenVINO. The highlight is a demonstration of this new feature in action. Check out the recording on YouTube:
Concerns about health, and an ever-present awareness of social distancing, is posing a significant challenge to the comfort levels once associated with touch-first digital content. More than ever, now is the time to explore touch alternatives, pursuing multi-mode deployments offering touch as an option, but providing additional options as well.
Computer vision is one of those alternatives. Nicknamed CV, computer vision enables applications to emulate human vision, accomplished through the application of deep learning inference engines supplied with pre-trained, optimized models of real-world phenomena. CV can be applied to just about any domain, but in the world of digital signage and other physical space deployments of digital content, CV is best known for face detection.
Thanks to its integration with the Intel® Distribution of the OpenVINO™ toolkit, Intuiface can now offer face detection to all of its users. Without writing one line of code, any Intuiface experience can be notified about the age range, gender, head pose, dwell time, and emotional state of each person in front of a screen. Acquired through any camera on a Windows PC, this data feed can be used to automatically curate content in real-time based on preset user criteria, and to enrich data sets for analytics with demographic information. And all information is anonymously collected, so no identifying information is in the collected data set.
The entire face detection solution is free to use.
Example use cases include:
- Prefiltering sale items based on age range or gender
- Resetting experiences after the detected user has moved away
- Enabling choice detection based on head position
- Combining demographic information with additional contextual info for retail analytics
- Responding to perceived emotional states with sympathetic messaging
- Playing music any time a viewer is detected across multiple cameras.
Here are two examples. The first highlights audience measurement to supplement analytics, the other highlights onscreen content manipulation with head movement.
The OpenVINO toolkit is open source, enabling Intuiface to offer the entire solution at no cost. It comes in two part:
- a "Face Detection with OpenVINO" Interface Asset, installed with Composer
- a Face Detection Server, the heart of the OpenVINO toolkit, that must be installed separately.
Installation of the Face Detection Server is a simple two step process - unzip a set of files on any Windows PC with a camera and then double-click a .bat file. (An Intel-based Windows PC is preferred as the OpenVINO toolkit has been optimized for used with Intel processors.) Deployments could vary. from 1) the same PC for the Face Detection Server and a running Intuiface experience, 2) separate Windows PCs for each, or 3) multiple servers and experiences communicating with one another.
Everything you need to know can be found in the Help Center article about face detection with OpenVINO.
The timing is fortunate, though the reason is not. We hope this latest feature from Intuiface inspires you to create experiences both innovative and sympathetic to an inevitable new normal.