Automated Video Meeting AI

One of the things we're interested in here at HTC are different approaches to building AI Bots. As we seem to be diving head first as a society into virtual meetings and telepresence as a normal part of daily life. Which immediately leads you to start thinking about how AI Bots are going to integrate or hijack themselves into your daily stream.

As a joke the other day i was mentioned how a certain government leader could be accurately emulated in an AI Bot by taking the repetitive speech patterns, rambling ever changing story line, etc, of their speech patterns, as well as a set of procedural animating image transformations to represent the visual part of the AI Bot. I was thinking of morphing transformations between different states in a random state machine.

Well, low and behold, i wake up the next morning and someone has already done it.  Matt Reed put together a digital twin to attend Zoom conference calls in his absence.

It was actually pretty straightforward.  He just recorded a few video shots of himself using the QuicktimePlayer (mac only, but you get the idea) and then saving them as a set of movie files for the AI Bot.

He then used ManyCam to setup a virtual webcam. The virtual webcam could then log into Zoom as a video conference attendee.

Reed had already been working with Artyom.js, a speech recognition and text-to-speech library, for another project. So, he knew it had the capability to listen for key phrases and speak back anything you program it to.

"It's basically a library to build your own custom voice assistant using standard web technologies," he says. "The best part is, you don't have to give it a wake word like 'Alexa' or 'Siri' to get it started. It listens for just the command phrases like 'How are you?' or 'Are you OK?' or 'Could you send that?'—which triggers a chain of commands like cycling through the stills of my face and speaking back the response text."

Want to build your own Zoombot? You can get a copy of it here and customize it.

Comments

Post a Comment

Popular posts from this blog

CycleGAN: a GAN architecture for learning unpaired image to image transformations

Image
 Like the Doublemint Twins touting the joys of Doublemint gum, 2 GANs are surely better than 1 GAN.  Especially if we package them together inside of one meta GAN module.  And this is exactly what the CycleGAN architecture does. Have you ever harbored dark secrets of turning a horse into a zebra?  The CycleGAN was developed to do just that. Learn how to turn a horse into a zebra.  And more. Now right away you can notice a difference between the image to image transformation GAN architectures we've been discussing over the last few posts.  Those last few posts described systems that learn from a database of matched input-output image pairs. And if your goal is to turn an edge representation into a nicely filled in continuous tone image, it's easy to build your database of matched input-output image pairs that your GAN system can then learn off of.  Take a continuous tone image (which will be the output of the database pair entry), then run it through an edge detector algorithm. 
Read more

Pix2Pix: a GAN architecture for image to image transformation

Image
 I thought following up yesterday's TraVelGAN post with a Pix2Pix GAN post would be useful to compare what is going on in the 2 architectures.  Two different approaches to the same problem. I stole this Pix2Pix Overview slide below from an excellent deeplearning.ai GAN course (note that they borrowed it from the original paper) because it gives you a good feel for what is going on inside of the Pix2Pix architecture.    Note how the Generator part is very much like an auto-encoder architecture, but rebuilt using the U-Net architecture features (based on skip-connections) that fastai has been discussing in their courses for several years before it became more widely known to the deep learning community at large  (and which originally came from an obscure medical image segmentation paper) . So the Generator in this Pix2Pix GAN is really pretty sophisticated, consisting of a whole image to image auto-encoder network with U-Net skip connections to generate better image quality at highe
Read more

Smart Fabrics

Image
"What makes smart fabrics revolutionary is that they have the ability to do many things that traditional fabrics cannot, including communicate, transform, conduct energy and even grow."  This is a quote from Rebeccah Pailes-Friedman, who is a professor at Pratt with a research focus on developing wearable technology and smart textiles. Enhancing textiles can be viewed through different lens.  Enhancements can be aesthetic, or performance boosting, or durability boosting. 
Read more