Knowledge Base

Detailed Log

The Detailed log contains all the expression classifier outputs and the other data you selected for export. Each expression has a value between 0 and 1, indicating the intensity of the expression. '0' means that the expression is not visible, '1' means that it is fully present. You can export the Detailed logs as text files with the data separated by tabs or as Excel files.

When no face is found or the face cannot be modeled (e.g. because the quality of the video is poor), a record with FIND_FAILED or FIT_FAILED is added. If you work live (using a camera) and your computer cannot process the incoming images fast enough, frames will be skipped and a record with MISSING will be added to the Detailed log.

NOTE To change the word 'Unknown' in the State log, see Export (Detailed log and API) in Settings.

NOTE If you want to use other criteria to define a state, for example when the intensity of an emotion is above 0.5, export the detailed log. Then calculate the states in another program, like Excel.

NOTE To change the words FIND_FAILED, FIT_FAILED, or Unknown in the Detailed log, see Export (Detailed log and API) in Settings.

If you analyze a video file and skip frames (that is, you set the Sample rate to Every 2nd frame or Every 3rd frame), the logging frequency changes accordingly.

For the Detailed log you have the following options:

• Adjust the export sample rate – To create a log file with a lower sample rate than the FaceReader analysis. You can choose a value between 1 and 15 frames per second. FaceReader interpolates the analysis results if the chosen sample rate results in interpolated timestamps.

  You can also choose to adjust the export sample rate to a value between 1 and 60 frames per minute. Note that this is averaging the results towards the new time base, centered at the time logged. So if you log 1 frame per minute, the result at the 1st minute will be the average of 0:00:30.000 to 0:01:30.000.

• Include file header – To include the following information as header in the export file:
  • Face Model
  • Calibration – calibration name and calibration values (the average intensity values of Neutral, Happy, Sad, etc. that are found during the calibration, that is, I_c in the formula used to calculate the corrected expression intensities. See How participant calibration works).
  • Start date and time
  • Video filename
  • Analysis frame rate
  • Independent variables
• Include column header – To include column details in the export file.

Note

If you analyzed from video and your video was not recorded with exactly the frame rate you selected, for example 29.9 fps instead of 30 fps, the timestamps may be slightly different from what you may have expected.

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