Stanford Startup Uses Face Tracking Technology to Make Training More Engaging

Blended learning and flipped classrooms have been hailed as the catchall solution and intermediary between traditional classroom learning and virtual learning settings.

Five years ago, Daphne Koller, a computer science professor at Stanford and co-founder of the EdTech company Coursera, realized that she would get up in front of students and lecture them on the same content, reciting the same stories and anecdotes. She said she wasn't forced to engage with students in a meaningful way.

As a solution, Koller helped pioneer the flipped classroom model, where instructors provide students with short YouTube videos to watch in between classes, and then structure in-person interactions as a forum for the big picture ideas and topics.

As freshmen at Stanford, Catalin Voss and Jonathan Yan found themselves in one of these flipped classroom settings for an introductory engineering course. Even with the effort to revolutionize the learning environment, Voss and Yan found they struggled to remain engaged.

"We just couldn't do it," Voss said. "Those 20-minute videos on YouTube were so boring that I would fall asleep five minutes in, and the answers to all of the quiz questions were all online."

The duo thought there must be a better way to leverage the technology available to both create more engaging content and measure the level of engagement among students.

Their brainchild: Sension, an EdTech solution that employs the ocular camera found on most smart phones, tablets and computers as a face tracking device to measure engagement.

Using this face tracking technology, the lecture videos on YouTube would pause when students look away from the screen. Then, when they return their gaze on the video, they would be prompted with a quiz question to assess comprehension.

State-of-the-art face tracking technology measures 76 points—around the eyes, nose, mouth and chin line—on the user's face. To measure all 76 points on a device like an iPhone or iPad proved prohibitively expensive, Voss said.

Sension's founders knew, however, that previous research found 95 percent of shape differentiators in people's faces—smiling, frowning, looking away, even putting their hands in front of their face—can be expressed in 17 modes.

Aside from higher education, this kind of smart technology that is reactionary to the user has clear implications for corporate learning and training.

"When things were mostly classroom based, instructors would know if students thought training was terrible or unnecessary," Voss said. "With the move to e-learning, educators get less real-time feedback about what parts of their content is effective."

The face tracking technology measures users level of engagement with every piece of content and then supplies the instructors with charts and graphs that detail where engagement dropped off and where it was highest. Sension provides an average of the engagement among all users and anonymizes all data, so it can't be used to punish individual students for their lack of engagement.

"Not only will these instructors know how effective their training is," said Mike Shore, an MBA student at Stanford and Sension's director of business development. "They can also make their training a whole lot better."

Sension has already partnered with a handful of corporate training companies including MindFlash. Voss said the company plans to make Sension available to a wider market within the next nine months.