Google have a new product, Google AR: Augmented Reality. They are selling it as an app which you can use to place virtual items in your home.
In the product’s web site they say:
…the possibilities of this technology are limited only by your imagination.
[…] many people have created amazing 3D models of animals and even real-life humans. We’re thrilled to see them shared and we’d love to see more.
[…] we’ve also seen some very cool AR experiences like AR golf and AR cooking.
[…] For now, the app only works on smartphones running Android 4.3 or higher with gyroscope support (which means an iPad 2 with a gyroscope is not supported). On iOS, the app is not yet available, but it will be soon through the App Store’s beta testing program.
We had been wondering about how to create an animal model for Google AR, and so this was an obvious opportunity to try it out!
In a few days, we will be showing off Google’s AR (augmented reality) technology at the IFA trade show in Berlin.
It’s a cool technology that you can use to place animals in your own home. You can also make them walk around or play tricks on people.
The design team have done a really good job of making it look like something magical is happening right before your eyes. It is almost impossible to tell which image is real and which one is not. The idea is that you can use this to change the way children play with pets and other toys in the home.
When you are standing there looking at it, it looks like magic, but when you sit down and someone asks what you are doing, they will see exactly what you have done – a cat walks into the room and then sits there next to them on their sofa.
The technology itself is fairly straightforward. The trick is how to make it look real when it isn’t.
The Google AR project is a very interesting and immersive experience, but I think it is not too useful for the average person.
The Google AR project uses advanced computer vision, machine learning and computer graphics to let you place objects in your house virtually.
It consists of two sets of software: one for the camera and one for the computer vision engine. The camera takes photos at a resolution of 2048×2048 pixels with a 3D depth map of the environment. The computer vision engine runs on the cloud in real time and learns how to recognize objects and select them from a database.
The database contains thousands of items that are downloaded automatically by scanning a QR code on the user’s phone or tablet. This information is then used by the computer vision engine to improve recognition capabilities, in combination with training data from the various other projects in which Google has invested in this area.
There is a tendency to think that Google’s technology is just for finding old-fashioned photos. This isn’t true. There are some pretty amazing things you can do with your phone that would not have been possible even 5 years ago, and you can use it in creative ways that wouldn’t have occurred to you before.
But it’s important to understand that there are limits to what the technology can accomplish. We’ve seen some examples of this in the news lately, so let’s look at two examples:
First, the BBC has covered a story about how they were able to take a video of a kitten playing at the bottom of a sink, and then use Google AR to bring the kitten into their studio. This is an excellent example of using the technology to make something more fun or more engaging than it would be without it. But what if you didn’t want to see a cat playing? What if there was something else happening in the background, like an emergency siren or something? Well, I don’t know about you, but I wouldn’t be able to turn it off. Not only would I have missed out on seeing the kitten, I’d have missed out on something else too.
Now imagine this instead: A man has his dog outside his house
The two main obstacles to creating a Google AR are the same as for any new technology: how do you create something that is useful and also going to be popular?
The first obstacle is technical. It’s hard to build a system that can recognize a cat, and then can display it in an environment where you can move around and see it from different angles. The second is that we don’t really know what kind of things people will want to use Google AR for.
Google has been working on a way to make 3D computer graphics for a while. You can already do it with Google Earth, and it’s easy, so why not?
For the past few years, Google has been working on it more seriously. It’s now working on something called Project Tango, which is probably what you call it when you can only explain it by saying “it’s like X but even cooler than X.” It’s harder to explain what AR means than what 3D graphics mean because there are so many things that are possible with 3D graphics today, but there are far fewer possibilities with AR.
AR isn’t just about putting 3D stuff into your house. It’s a lot more interesting. Imagine buying a cat and then being able to see what the cat would look like in your room at any time. Or being able to buy a computer that comes with a custom-printed display that shows you what objects would look like if you placed them in your house.
This is a blog post about how to use Google Glass wearers’ heads to create an interactive 3D model of some animal. For example, you could take a picture of a bird and see it emerge in your living room as a lifelike bird with moving wings.
The post is also a sort of virtual reality demo. I walk through the process of putting together the model, which involves taking many pictures from different angles and stitching them together into one high-resolution image that looks like it came out of a photo printer. The process is real enough that if you turned off your phone’s camera, the image would capture your whole living room exactly as you see it now. It would be totally indistinguishable from reality except for the fact that you are seeing it via Google Glass.
My interest in making this kind of thing started with my experience at TEDxSF, where I talked about how we can use computer vision to create something called “augmented reality.” In this case, augmented reality means turning real objects into computer-generated ones without destroying their physicality. We could show people things they can touch and make them visible on their phone screens but also have the ability to manipulate them through small motions using the device’s built-in controller.
To take this