Amsterdam Interdisciplinary Centre for Emotion (AICE)

Stimulus Set Contents

The Amsterdam Dynamic Facial Expression Set (ADFES)

The files contain the full video stimulus set in MPEG-2 formats, plus some additional High Dynamic (HD) recordings and still pictures created from the stimuli. If you would like to receive other formats, or other still pictures, you can contact us.

Expressions

The set is comprised of displays that represent nine emotional states (anger, contempt, disgust, embarrassment, fear, joy, pride, sadness, surprise) and “neutral”.

Models

Twenty-two models (10 female, 12 male; 10 Mediterranean, 12 North-European) feature in the set, all between the ages of 18 and 25. Most models came from either North European or Mediterranean (Turkish or Moroccan) ancestry. Further information about the specific ethnic background of the models is available upon request.

Directedness of expression

Three versions of the displays were created (face-forward, turn-toward, and turn-away). The face-forward version involved models beginning with a neutral face, and then expressing the emotion directly to a camera at a face-forward (0°) angle. The turn-toward and turn-away versions involved filming the model from two angles, simultaneously. Models began by directly facing the first camera (0°), with a neutral expression, and then turning to face the second camera (45°, on the right) before making the facial expression. This created two versions of the same film – one with the model turning toward the viewer before making the relevant expression (the turn-toward clips), and the other with the model turning away from the viewer before making the expression (the turn-away clips).

Timing of the stimuli

All films include a neutral face for .5 seconds, followed by the onset of the expression, and then the face held at apex for 5 seconds. The length of the turn-toward/-away films is slightly longer, in order to accommodate the head turn. The films thus range from 6 to 6.5 seconds.

Missed recordings

Due to time constraints, we failed to record embarrassment, pride, and surprise expressions for one of the female Mediterranean models (F10), and pride expressions for one of the male Mediterranean models (M1).

The set is therefore comprised of 10 (displays) x 22 (models) - 4 (missed recordings) = 216 stimuli per head turn version, 648 stimuli in total.

Extra recordings

“High Dynamic” recordings
“High dynamic” (HD) video clips (with extra face, head, or body movements) of contempt and embarrassment are available for most North-European models (except for the turn-toward/away versions of contempt for F2), and some Mediterranean models. Most notably, the embarrassment HD stimuli include the model touching their face/hair (see Keltner, 1995), and the HD contempt pictures feature a head tilt and/or a unilateral eyebrow raise. These clips are denoted with the suffix “(HD)” following the emotion label.

Also included is an extra recording of joy for female Mediterranean model F10 with open mouth (AU25).

Still pictures
Still pictures (in JPEG-format) were produced from the face-forward versions by freezing the film at the apex (point of most intense muscular contraction) of the model’s expression. When the apex expression differed slightly from the targeted facial prototypes, we also created more prototypical pictures. These additional pictures can be found in a separate folder within the “Still Images” menu and are denoted with the suffix “PR”.

Standard Facial Expression Codes

The faces used for the ADFES are based upon prototypes of the “basic emotions” defined in the Facial Action Coding System (FACS) Investigator’s Guide by Ekman, Frisen, & Hagar (2002). Displays of embarrassment are based on the prototype identified by Keltner (1995), and displays of pride are based on the prototype defined by Tracy & Robins (2004). Two coaches trained in FACS worked with each model in order to elicit these expressions through a variation of the Directed Facial Action Task (DFAT; e.g., Ekman, 2007). Below are the target codes (in terms of FACS Action Units) that we attempted to elicit from the models. While some additional movement was allowed, action units in boldface are "core units" that were considered to be absolutely essential, and were the main requisites for selecting stimuli used in initial pilot tests.

Anger: 4CDE + 5CDE + 7 + 17 + 23/24
Contempt: U1 + U2 + U14
Disgust: 9 + 10 + 25
Embarrassment: 12 + 14 + 23/24 + 54 + 64
Fear: 1 + 2 + 4 + 5DE + 20 + 25
Joy: 6 + 12CDE + 25
Pride: 6 + 12 + 53 + Posture straighten
Sadness: 1 + 4 + 15ABC + 17
Surprise: 1 + 2 + 5AB + 26

References

  • Ekman, P. ( 2007). The directed facial action task. In J. A. Coan and J. J. B. Allen (Eds.), Handbook of Emotion Elicitation and Assessment (pp. 47-53). Oxford University Press.
  • Ekman, P., & Friesen, W. V., & Hager, J. C. (2002). Facial action coding system: The manual on CD-ROM. Instructor's Guide. Salt Lake City: Network Information Research Co.
  • Keltner, D. (1995). Signs of appeasement: Evidence for the distinct displays of embarrassment, amusement, and shame. Journal of Personality and Social Psychology, 68, 441-454.
  • Tracy, J. L., & Robins, R. W. (2004). Show your pride: Evidence for a discrete emotion expression. Psychological Science, 15, 194-197.

Published by  PRI

31 August 2015