Monday, November 7, 2016
Interview with Nikhil Jain, ObEN
Story by Benjamin F. Kuo
What if you could—with your smartphone—easily create a virtual you—an avatar that looks and sounds like you, and which would live in virtual or augmented reality? That's the idea behind the technology being developed by Pasadena-based ObEN (www.oben.me), which is led by co-founde rand CEO Nikhil Jain. Nikhil told us a bit about the startup,which was spawned out of Idealab, and recently raised a round of funding from CrestValue Capital, Cybernaut Westlake Partners, Leaguer Venture Investment Co., Ltd.,, Third Wave Digital (Allen DeBevoise), Dream Maker Entertainment Ltd. (S.M. Entertainment Co., Ltd.), Gordon Cheng (Cameron Pace Group China), and Idealab.
Tell us a bit about what your startup is working on?
Nikhil Jain: We are creating, for the first time, the ability for any person to combine their image and their voice to create an amazing, realistic, and personalized 3D virtual self, which can be transported into an Augmented Reality or Virtual Reality experience, and made social. The way it works, is you take a selfie with iOS or Android, and without extra hardware or attachments or scanning, you can do a voice recording, and using our proprietary artificial intelligence technology, we automatically create a photo-real, virtual self of that person. The goal has to make the technology really accessible using the power of the smartphone, and bringing people into AR and VR as no one has done before in such a way.
How did you come up with the idea for the company?
Nikhil Jain: My co-founder and I met at Idealab. We were exploring different ideas on what we wanted to do next. We were looking at ways of connecting technology in an easy manner to consumers. I was traveling one time for work. I have young kids at home, and my wife called me, because she was frustrated she could not put our kids to sleep. One of the reasons for that, was my kids were so used to me telling them a story every night, they just did not want to go to sleep. The reason for that, was my kids were so used to me telling them a story every night, and they just did not want to go to sleep. I read them the story over the phone, and I shared this idea with my co-founder. What we came up with, was a way to create a virtual self, while I was traveling, that could read the storybook to my kids, which would look like me, talks in my own voice, and my kids are happy. That's the entire premise of the idea.
Where is the product now?
Nikhil Jain: We plan on doing our product launch in Q2 of 2017. At this time, we're working with lots of our partners for the launch. It's not accessible to consumers at this time. Our goal is to announce the Series A, and give a quick preview of the company and what is all about, and we will be coming out of stealth for the first time.
Can you talk a little bit about your backers and why they're working with you?Nikhil Jain: We actually are working on different use cases, in areas of entertainment, healthcare, and enterprise, and also in gaming, which is the reason a lot of these partners starting working with us. The way our technology works, it's very easy to use. Traditionally, to create the types of avatars we are creating, you'd have to go into a studio and scanning booth, and scan in your information to make that avatar. The alternative, is a $400 depth sensor for your phone. We decided that just wasn't something consumers would use, so we came up with our own technology, which just uses a regular smartphone camera. A lot of our partners liked that, because the product we are making with them is going to available to the mass market, in an easy-to-use manner. That was how it started out. Although people generally associate AR and Vr with gaming, we're actually growing into a lot of other areas, like healthcare, education, and the enterprise.
What was your biggest challenge in building this?
Nikhil Jain: There are three different pieces to this. The first, is the ability for a consumer to take a selfie, which looks like you, just in a few seconds. The second piece, is to let you speak into a smartphone and make a voice print of you. This voice print can be applied to any IoT, or avatar of that person. Your IoT and visual avatar can speak in your voice, providing any content that you and I could speak. The third piece of the technology, combines the first two, in a way that animates the mouth of your visual avatar. Say your avatar voiceprint speaks a particular sentence, your visual avatar has to move with the right kinds of gestures and muscle control to reflect what they are speaking. Those are the three elements to our technology. Everything was developed in-house at Idealab. We've already filed fifteen patents already on our technology, but it's definitely been a challenging couple of year sto come up with all of this and to find the talent to carry it out. We have about 25 people in house, and around 15 outside whoh have worked on this. We've been hiring from all of the top schools—Carnegie Mellon, Caltech, USC, Berkeley, UCLA ,and it has taken a combination of deep learning, machine learning, speech synthesis, linguistics, and psychology to come up with all of this.
What has been the biggest lesson you've learned from your past startups which you are applying here?
Nikhil Jain: Perseverance. That's a huge one. With this kind of startup and technology, people kept saying to us, making this that accessible to consumers is not going to work. But, we were very adamant that this is the only way consumers will use this technology. The ability to stay in the game, and focus on the fundamentals is key. The second thing, is you have to listen to consumers. You have to understand what they want, what their problems are. We realized the problem was that people wanted to use avatars, but have been stuck with avatars that look like cartoon characters, or are just the default characters. We learned early on that we needed to address that problem. Third, you have to invest heavily into the best team you can, and come up with a business model which is sustainable for you, your partners, and developers.