This FAQ provides information about the hardware requirements and features included in the Intel® RealSense™ SDK Gold R1 and R2.
Requirements
- What are the hardware requirements to run and develop with the Intel® RealSense™ SDK?
- 4th gen Intel® Core™ processor (code named Haswell) or later
- 8 GB of free hard drive space
- The Intel RealSense 3D camera
- What are the software requirements?
- Microsoft Windows* 8.1
- Microsoft Visual Studio* C++ 2010 ~ 2013 with service pack 1 or newer
- Microsoft .NET 4.0 or newer framework for C# development
- Unity* PRO 4.1.0 or later for Unity game development
- Processing* 2.1.2 or later for Processing framework development
- openFrameworks* v0071 or later for openFrameworks development
- Havok Vision* SDK 2012.2.1 or later for Havok Vision SDK development
- Java* JDK 1.7.0_11 or higher for Java development
- Any of the following browsers for JavaScript* development:
- Microsoft Internet Explorer 10.0.13
- Google Chrome* 33.0.1750.146
- Mozilla Firefox* 27.0.1
- Does the Intel RealSense SDK work with any other camera?
- No. It only works with the Intel RealSense front facing camera.
- Is there a Linux* version of the SDK?
- No.
Performance
- What is the CPU utilization on the Intel RealSense SDK compared to the Intel® Perceptual Computing SDK?
- 30% for a single module and < 50% for combined modules. This is a rough guideline and performance may vary.
Unity
- What versions of Unity are supported?
- Unity v4.1 Professional and above
- What is the Unity Toolkit?
- The Unity Toolkit is an Intel-developed Unity package that aids Unity game developers by allowing for interaction with the Unity Inspector drag-and-drop type operations. The toolkit functionality replaces a lot of the boiler plate coding that had to be done when using the Intel Perceptual Computing SDK.
- Are there any requirements for using the Unity Web Player?
- No permissions are needed.
- Are there any Unity limitations in the Intel RealSense SDK like there were in the Intel Perceptual Computing SDK?
- No, all features that are available in the SDK should also be available to Unity developers.
Facial Recognition
- How many faces can be detected at one time?
- Four
- How many faces will have landmark points at one time?
- One
- What head orientations are tracked?
- Pitch, Roll, Yaw
- How many landmark points are there?
- 78
- What are the regions of the face that landmark tracking is supporting?
- Eyes, eyebrows, nose, mouth, and jaw
- Does the Intel RealSense SDK support facial expressions and if so, what are they?
- Yes, the SDK does support the following facial expressions:
- Smile
- Eye-brow raised
- Eye-brow lowered
- Mouth Open
- Eyes closed
- Pupil move
- Does the Intel RealSense SDK support facial emotion and if so, what are they?
- Yes, while still experimental, the SDK does support the following facial emotions:
Expression R1 R2 Anger Disgust Fear Joy Sadness Surprise Contempt Positive Negative Neutral
Gestures
- What gestures are supported in the beta SDK?
- The following are the current gestures supported by the SDK:
Gesture R1 R2 Fingers Spread V Sign Tap Wave Two Finger Pinch Open Fist Thumbs Up Thumbs Down Click Swipe Up Swipe Down Swipe Left Swipe Right - Will gesture recognition support features like holding a pen?
- There is no support for this type of gesture.
- What is the target date for the Custom Gesture Tool?
- Target date is in 2015. The exact timeframe has not yet been determined.
Speech
- What type of speech modes does the Intel RealSense SDK support?
- Command Mode and Dictation mode
Miscellaneous
- As I understand it, there is no backward compatibility between the Intel RealSense SDK and the Intel Perceptual Computing SDK. Does this mean apps will have to be re-written to be used with the new camera and the 2014 SDK?
- Applications based on the former SDK must be adapted to the new one. While the general concept remains the same, there have been numerous architectural changes in the new version which prevent compatibility.
- Can multiple apps get data from the same camera device? And if so, how does the middleware handle the different resolution settings that two different apps might want to use?
- Two apps can get data from the same camera device if they use the exact same settings. If not the same settings, the camera will function on a first come, first served basis. Other applications requesting a different resolution will get an error.
- If App-A changes a setting, will App-B get some kind of notification about the setting change?
- App-B can explicitly monitor the device property changes.
- Can the device be queried for power state? (Is the camera on? Is the camera off?)
- Yes.
- Can a slave app query the properties set by the master application?
- Yes.
- Must special device drivers be downloaded separately to support the new integrated camera?
- Camera drivers and firmware are available from the OEMs. For discrete development cameras, use the firmware and drivers from the Creative Labs Intel RealSense support website.