Kinect for Windows – WebProNews

Tag: Kinect for Windows

You Can Now Apply To Pay $400 For The New Kinect For Windows Sensor

Microsoft is really onto something with the Kinect. Sure, it’s not that great for games, but developers are doing some amazing things with it in the application space. It’s only going to get more amazing next year as the new and improved Kinect comes to PC.

At its annual BUILD developer conference, Microsoft announced that it’s opening up applications for preorders for the new Kinect for Windows sensor. It’s the very same Kinect that will ship with the Xbox One later this year so it has the improved camera that can reportedly detect heartbeats.

The new Kinect for Windows sensor won’t be available until next year, but interested developers can get their hands on one this year. To do so, you must apply to take part in the Kinect for Windows developer kit program. If you are accepted, you’ll be given a chance to pay Microsoft $400 for the following:
Direct access to the Kinect for Windows engineering team via a private forum and exclusive webcasts
Early SDK access (alpha, beta, and any updates along the way to release)
Private access to all API and sample documentation
A pre-release/alpha sensor
A final, released sensor at launch

You could argue that $400 is a little too much for this, but you are getting two Kinect sensors. Granted, we don’t know how much the new Kinect will cost separately as Microsoft is packing one into every Xbox One sold. Those retail for $500, and the original Kinect retails for $150, so it stands to reason that the new Kinect will probably retail for $200. In short, you’re paying for two Kinect sensors – one for this year to start developing for and a final unit for next year.

If the idea of developing for the next generation Kinect appeals to you, you can apply to the Kinect for Windows development kit program here. All applications must be turned in by July 31 and Microsoft will notify successful applicants in August. Those were accepted can expect to get their alpha Kinect sensor in November.

Kinect Code Samples Go Open Source To Encourage Development

Kinect for Windows has inspired developers to create some really impressive applications. That’s apparently not good enough, though, as Microsoft is now making Kinect for Windows partially open source to encourage more development on the platform.

The Kinect for Windows team that its sample code is now open source under an Apache 2.0 license. The 22 samples are available in C#, C++ and Visual Basic.

So, why go open source? The team says that there are four reasons for the move:

Kinect Fusion Brings 3D Modeling To Kinect For Windows

Kinect For Windows is one of Microsoft’s best projects to date. It’s a bit gimmicky and not that great for games, but it’s been amazing for app developers on the PC. Back in September, a developer created an email system for his mother who suffered from aphasia after a stroke. Now Microsoft is adding a new tool that will inspire a whole new range of applications.

At BUILD 2012, Microsoft announced that Kinect Fusion will be coming to Kinect For Windows. Kinect Fusion is software that creates a 3D model of an object, or even the environment, in the Kinect sensor’s view. The tool can be used for any number of applications, from augmented reality to 3D printing.

So how does it all work? Chris White, Senior Program Manager on Kinect for Windows explains:

Kinect Fusion takes the incoming depth data from the Kinect for Windows sensor and uses the sequence of frames to build a highly detailed 3-D map of objects or environments. The tool then averages the readings over hundreds or thousands of frames to achieve more detail than would be possible from just one reading. This allows Kinect Fusion to gather and incorporate data not viewable from any single view point. Among other things, it enables 3-D object model reconstruction, 3-D augmented reality, and 3-D measurements.

The team making Kinect Fusion uploaded a video back in August that shows off the tech. As far as Kinect goes, it’s really impressive.

Kinect Fusion will be available in a future release of the Kinect for Windows SDK. It will be interesting to see what new applications developers will build with the technology.

[Image Source]

Kinect For Windows SDK 1.5 Released

Kinect for Windows has been one of the coolest projects that Microsoft is personally involved with. Windows applications being built using Kinect continually impress me more so than any games that use Kinect on the Xbox 360. It speaks to the power of the Kinect platform as an interface.

With that being said, Microsoft has updated the Kinect for Windows SDK and Runtime to version 1.5. The updates adds a host new capabilities and tools to make building Kinect-powered programs easier. The updates range from new tools like Kinect Studio to new languages in Kinect’s speech recognition.

Starting with the new tools, developers can now play around with the above mentioned Kinect Studio. It allows developers to record and play back Kinect data. This should make testing easier since you don’t have to have the person perform the movements over and over again.

Microsoft has released a set of Human Interface Guidelines that detail best practices when creating Natural User Interfaces with Kinect. On a related note, the Face Tracking SDK is now available. It tracks head position, location of eyebrows, shape of the mouth and more.

The skeletal tracking capabilities have been greatly enhanced with this release. The biggest update is the addition of Seated Skeletal Tracking. One of the major problems with Kinect on Xbox 360 is that it can’t detect movement very well when the person is sitting down. With Kinect for Windows, it can now detect a 10-joint head/shoulders/arms skeletons while ignoring the leg and hip joints.

Skeletal Tracking has also been added to Near Mode. It supports Default and Seated tracking modes. The idea here is that Kinect can be used to track movement in applications that involve displays which people need to interact with on a personal level.

Performance has seen improvements with the mapping of a depth frame being sped up to five times from the last update. The Kinect device will also keep depth and color frames in sync with each other. The RGB image quality has also been increased with 640×480 now running at 30 FPS and YUV 640×480 running at 15 FPS.

Adding in a feature from Kinect for Xbox 360, it’s now easier to build applications that allow users to control 3D avatars. It does this by providing Joint Orientation information for the skeletons being tracked. The Joint Orientation is offered in two flavors – Hierarchical Rotation based on bone relationship or Absolute Orientation using Kinect Camera coordinates.

Kinect is also becoming more internationally minded with Microsoft releasing four new language for speech recognition – French, Spanish, Italian and Japanese. They also released new language packs to support different dialects based on region. This is especially important for English as Kinect has to contend with differences between the U.S., Canada, UK, Ireland, Australia and New Zealand.

Finally, the Kinect for Windows hardware is now launching in Hong Kong, South Korea, Singapore and Taiwan. It will be available in 15 additional countries (mostly European) in the following months. Microsoft is a global company so I wouldn’t be surprised if they had Kinect for Windows hardware in every major country by the end of next year.

It really is exciting to see Kinect for Windows being supported like this. I was concerned that Microsoft would treat it as just a side project to their big money maker – Kinect for Xbox 360. I have been proven wrong, but it wouldn’t have been possible without the amazing support that Kinect for Windows has so far received. To see some of the amazing things people are doing with Kinect, check out the Kinect Accelerator finalists.

To download the latest SDK and Runtime, just hit up the Kinect Developer page. It’s easy and free. You can’t beat that combination.

Kinect For Windows Gets A Gesture Recognition SDK

It’s been said before and I’ll say it again – Kinect isn’t just for video games anymore. Ever since the release of the Kinect for Windows SDK last year and the official Kinect for Windows hardware this year, developers have gone to great lengths to create some of the most creative and awe inspiring applications we’ve seen in years. A few features have been missing but Carl Franklin has the answer to at least one of them.

Franklin has created a gesture recognition SDK for Kinect for Windows called GesturePak. With this new software, developers will be able to easily program gesture recognition into their applications. It would be especially useful for applications that require such recognition since building it yourself would take months.

Even if it did take you months, you could still build gesture recognition for yourself. That’s stupid though since Franklin already did the hard work for you. Here’s his explanation:

“When Microsoft released the Kinect for Windows SDK beta last year, I tried to write an app to recognize a simple gesture. It was way too complex. The SDK spits out a stream of joint data (X,Y, and Z axis data points for each of the 20 locations on your body that the Kinect tracks) at 30 frames per second. In order to recognize gestures you have to track coordinates in space over periods of time, compensate for a margin of error, and somehow determine that the user is actually moving deliberately the way you want them to.

My idea was to simplify this process by breaking down a gesture, say a hand wave, into poses. A gesture is a series of poses that are matched sequentially. So, I wrote a recorder (I’m an audio recording engineer so my brain just goes there…) that takes snapshots of your body and saves the data into a collection of poses. Then I wrote a matcher that watches you in real time, and determines if you’re hitting the poses. It’s a lot of logic and math, but at the end of the day all that goo is abstracted away from the programmer. If you want to recognize somebody flapping their arms, rocking their head from side to side, kicking their feet, swaying their hips, or just about anything you can think of, you’re crazy if you don’t use some kind of gesture matching toolkit like GesturePak.”

Franklin says that Microsoft themselves are working on gesture recognition solutions for version 1.5 of the Kinect for Windows SDK. The solutions Microsoft provides isn’t real gesture recognition, Franklin says its getting close. In the meantime, his SDK can take care of the work for developers wanting to track gestures.

Head over to the Web site to try out a free demo of the gesture recognition SDK. It allows you to “create and test as many gestures as you like, even matching multiple gestures at the same time. If you want to save those gestures, however, you’re going to need to buy the license. With the license, you also get the power to have your Windows Apps recognize your saved gestures.

A single developer license only costs $99 with a site license going for $799. Considering that a lot of the Kinect developers who won the Kinect Accelerator were small teams, this kind of pricing structure will be most advantageous to the kind of developers making the truly unique applications with Kinect.

Kinect For Windows Launches Today

Kinect for Windows launched today – pause for celebration.

Now that the celebration is over, we can get into the stuff that really matters on the Kinect for Windows front – software. Over on the Microsoft blog, the Kinect team detailed changes being made to the beta 2 SDK that will move the Kinect software up to version 1.0. As previously reported during CES, the retail price of Kinect for Windows is $250 with special academic pricing being set to $150.

The list of features added to version 1.0 really helps to set Kinect for Windows development apart from its Xbox 360 equivalent. People should be able to take the device to new highs with these new features. As an aside, it can be assumed that some of these features will be showing up in Kinect for Xbox 360 dev kits as well.

Support for up to four Kinect sensors plugged into the same computer.

Significantly improved skeletal tracking, including the ability for developers to control which user is being tracked by the sensor.

Near Mode for the new Kinect for Windows hardware, which enables the depth camera to see objects as close as 40 centimeters in front of the device.

Many API updates and enhancements in the managed and unmanaged runtimes.

The latest Microsoft Speech components (V11) are now included as part of the SDK and runtime installer.

Improved “far-talk” acoustic model that increases speech recognition accuracy.

New and updated samples, such as Kinect Explorer, which enables developers to explore the full capabilities of the sensor and SDK, including audio beam and sound source angles, color modes, depth modes, skeletal tracking, and motor controls.

A commercial-ready installer which can be included in an application’s set-up program, making it easy to install the Kinect for Windows runtime and driver components for end-user deployments.

Robustness improvements including driver stability, runtime fixes, and audio fixes.

Microsoft hopes to release updates to the SDK and runtime two to three times a year. The team is already working on the second update as you read this. They are also continuing to invest in other programs like the Testing and Adoption Program and the Kinect Accelerator. They will continue creating new programs in the future to support their developer and partner ecosystem.

With the amazing stuff people have already put out using the beta drivers, it will be interesting to see what developers do with custom-tailored for Windows software. It’s also safe to assume that whatever the Kinect for Windows team cooks up will make its way to a Kinect for Xbox 360 near you as well.

To see what Kinect for Windows can already do, have a video of Skyrim on PC being controlled via Kinect: