Knowledge Base

FaceReader Trial Version

Aim

To try FaceReader for free for a limited period.

Restrictions

• Valid for 14 days (from the day of activation).

Prerequisite

• You received a link to download the setup file.
• You received a software license key.
• Your computer meets the System requirements

Procedure

1. Open the FaceReader trial download page and download the setup file.
2. Double-click the setup file that you downloaded and follow the instructions on your screen.
3. Start FaceReader.
4. A window opens, choose Activate software license key.
5. In the following window select either Floating or Fixed. See What Is The Difference Between A Floating And A Fixed Activation?
6. In the window that appears, enter the software license key.
7. Click OK.

Notes

• To see how many days are left in your trial, choose Help > About.
• When the trial period expires, a message is shown at start up.
• The Action Unit Module, The Vital Signs Module, The Advanced Research Module, and The Consumption Behavior Module are enabled in the trial version. With the Trial version you can use the regular FaceReader functionality and Baby FaceReader.

Camera And Accessories

Supported Cameras

FaceReader can only classify facial expressions correctly if it gets a good image. Both the placement of the camera and the lighting of the subject's face are of crucial importance in obtaining reliable classification results. Also the quality of the camera is crucial. The following cameras are supported.

USB cameras

FaceReader is supported with current generation CCD webcams with a resolution of at least 640 x 480 pixels. FaceReader and all its modules has been tested with the following USB cameras:

• Logitech C920, C922, C925e, C930e
• Logitech Brio
• Microsoft Lifecam Studio

Do note that the Logitech Brio and Microsoft LifeCam Studio are not suitable for the Vital Signs Module due to internal image processing (see Introduction to Vital Signs).

You can purchase these cameras from us or use another current generation high quality webcam with a resolution of at least 640 x 480 pixels.

IP cameras

An IP camera is connected with the FaceReader PC via Internet, and enables you to do tests with test participants at another location or at the test participant's home.

FaceReader works with IP camera that support MJPEG or JPEG output. FaceReader cannot import RTSP output of IP cameras.

Compared to webcams such as the Microsoft LifeCam Studio, IP camera signals are processed slower. This is caused by the fact that IP cameras need time to encode video. Also sending videos over a network takes time. This means that the IP camera video signals received by FaceReader are analyzed with a delay. This effect is also seen when The Stimulus Presentation Tool is used. The delay is in the order of magnitude of 0.1 - 0.3 seconds if you use the FaceReader setup in a company network. If you use a dedicated local network for the FaceReader setup, which is recommended, the delay is less and in the order of magnitude of 0.1-0.2 seconds.

If this delay is too large for your experiment's accuracy, record with a USB camera instead.

Infrared cameras

You can also use a camera that is sensitive to (near) infrared light, in combination with lights of this type, for instance, if you want to analyze people's facial expressions in a cinema or other (near)-dark environments. Please note that such cameras are not suitable for vital signs measurements and it is not possible to measure heart rate, heart rate variability, and breathing rate with infrared recordings.

To get started with the Microsoft LifeCam Studio

Download the drivers (Microsoft LifeCam 5.20.msi) from the MyNoldus portal (my.noldus.com).

IMPORTANT If you upgrade an older version of the Microsoft LifeCam Studio driver to the driver that comes with FaceReader 10, you cannot downgrade it to the old driver anymore.

To get started with the Logitech BRIO, C920, C922, C925e and C930e

The Logitech BRIO, C920, C922, C925ed and C930e do not require driver installation.

See also

• Camera Frame Rate And Samples In FaceReader
• Camera Resolution And Analysis In FaceReader

Camera Position

It is important to place the camera at a location that will give the most steady frontal view of the subject's face throughout the experiment. The ideal position for the camera is directly in front of the test participant. If the subject faces a computer screen, the camera can be placed either directly above or directly below the screen. When placing the camera on top of the monitor, try to position it slightly below eye level, either by raising the chair and/or by lowering the monitor. When placing the camera below the monitor, lower the chair to position the camera slightly below eye level. The classification output may have a small bias towards 'angry' when the camera is placed on top of the monitor and a small bias towards 'surprised' when the camera is placed below the monitor. This is due to the fact that people tend to tilt their head when showing these emotions. If that is a problem, you can compensate for it by calibration.

See Calibrate FaceReader

Image Brightness And Contrast

It is important to adjust your camera to provide images with a good contrast and brightness. You can adjust the contrast and brightness by adjusting the hardware settings using the software provided by the manufacturer of your camera. When adjusting your hardware settings, do not pay attention to the lighting of the background. Focus only on the face area of the image and ignore a very light or very dark background.

Do note that for the Vital Signs Module, slightly different camera capture settings apply. See Introduction to Vital Signs for more information.

The picture above is an example of a face image showing a good contrast and lighting. Dark areas of the face (such as the eyebrows and the pupils) are near-black and the lightest part of the face (the eye whites) are nearly white. The intensity histogram (made of the face area only) shows how the intensity values fill the whole spectrum.

The face area of the image below is overexposed. The forehead and cheekbones are nearly white and show very little texture.

The histogram shows how only the high intensity part of the spectrum is filled. Decrease the brightness by adjusting the hardware settings of your digital camera or decrease the aperture size of your analog camera to correct this problem.

A very common problem is that the image has too little contrast.

Especially when the background is very dark or very light and the camera is set to 'automatic', the face area of the image will have a very low variability in intensity (a low contrast). The contrast is too low when the face area of the image contains no bright white and no deep black pixels. The intensity histogram shows that only the middle section of the spectrum is filled. Correct this problem by adjusting your camera (increase the contrast).

When working with a person with a dark or very pale skin color, you may need to make some extra adjustments. The default settings will often yield an image as shown below. Although there are both dark areas and white areas (the eyes) in the face, the skin itself shows too little variations in intensity. By increasing the brightness (to make the face lighter) and increasing the contrast (to increase the variation in skin tones) you can correct this problem. For people with a very pale skin, decrease the brightness and contrast.

The corrected image should look similar to the image below. Pay special attention to the skin tones.

Notes

• IMPORTANT Changing the brightness and/or contrast of an already recorded video or an image will have little or no effect on the quality of the analysis. The procedure above can only improve FaceReader's performance when applied at the moment the video is made.
• It is not possible to view intensity histograms in FaceReader. To view such plots, use graphics software like Paint.NET or Photoshop. To analyze video images, first make a screenshot of the image of interest.
• To assess the quality of your (video) image, view FaceReader's model quality bar.
  See one of the following:
  - Step 3 in Manual Procedure Image Analysis in Analyze Facial Expressions
  - Step 3 in Manual Procedure Video And Camera Analysis in Analyze Facial Expressions

See also

• Camera Frame Rate And Samples In FaceReader
• Camera Resolution And Analysis In FaceReader

Lighting Setup

A good lighting setup is vital to get a good image. Diffuse frontal and natural lighting is desirable. The light intensity or color is less relevant. Strong shadows or reflections should be avoided. If possible, place the FaceReader setup in front of a window. Make sure that any windows to the sides of the subject are blinded. Lights from the ceiling, common in most buildings, will produce shadows below the eyebrows and nose. In situations where interior lighting cannot be controlled, stronger lights (e.g., professional photo lights) can be used to negate the effect of other, undesirable, light sources. You can purchase such lights from Noldus IT. You can also use (near) infrared light, in combination with a camera that is sensitive to this type of light.

Source: FaceReader 10 Reference Manual (Help), Noldus Information Technology