Throughout the history of art production the self-portrait has existed.  Through its practice humanity can indulge its enduring desire for self examination and identity defining.  By definition, the practice requires investigative looking at one's self, staring into a mirror to accurately render the physical characteristics.  Equally integral is the representation of the self in the context of society resulting in the expression of personal values, professional and/or social status and qualities of character.  Resulting images in painting often place a dignified figure holding the tools of the trade, brush and palette.   These images were often used as advertisements to show a successful, accomplished and skillful artisan.  In modern and post-modern art, these self-portraits have become more personal and psychological (Frida Kahlo, Salvador Dali, Leonora Carrington, Connie Imboden ).  Other artists working primarily in self-portraiture produce work critical of contemporary society and the history of art (Cindy Sherman, Kim Dingle, Robert Arneson)

In my project, the AutoPortrait alludes to the tradition of self-portraiture while in name suggesting the culture of automation. In all cases, the self, is the medium for creating the self portrait.  Through sensors developed to recognize emotion and a digitally photographed image of an individual, an AutoPortrait is produced.  It is instantaneous and an accurate reading of the self at the moment of portrayal.
 The idea of the accuracy of the reading is, of course, arguable.  The sensors are after all producing empirical data, but the translation is a malleable judgment.  Yet, the
of person or machine are both equally flawed.

The AutoPortrait Project and Technology

Viewers of the AutoPortrait project will walk up to a black screen framed in ornate gold leaf.  The viewer will be attached to sensors with which they can accurately examine themselves for self portraiture.  The sensors monitoring bio-signals include a GSR (Galvanic Skin Response),  EMG (Electromyogram, BVP (Blood Volume Pressure)m Respiration Rate and temperature.  These sensors have been developed at the MIT Media Lab


These sensors will establish the emotional state for the AutoPortrait. Next, a digital image will be established of the viewers facial features along with a "soft mask" of their neutral expression through thorough measurement.  This "soft mask" is based on the research done for direct face to face communication between real and virtual humans published as an article through the MIRALab at the University of Geneva in 1998.


The research done here, specifically, on the digital reproduction of emotion onto a virtual person's face is applied to the AutoPortrait project.  However, the image established, in this case, is from the features of a specific person and the input of emotion from sensors monitoring bio-signals.

Since this research creates a virtual human presence, the AutoPortrait will have a life-like appearance.  Simple examples would include a "sad" sensor reading resulting in tears continuously running down the face of the AutoPortrait which simulates slow deep breathing. This would create a rounded life-like image and allude to religious miracles of the weeping virgin.  An anxious or nervous sensor reading would result in deep forced breathing and a furrowed brow.  Alluding to a tradition in painted or photographed portraiture, a happy or peaceful sensor reading would result in a background of  nature represented with full light and blooming natural life.

The result of this is a self-portrait, or, AutoPortrait, of the individual viewer. After it's completion and the departure of the viewer, the screen will return to black.  If the same viewer revisits the same  AutoPortrait unit,  technology from the same MIRALab research on facial feature recognition based on skin hue and edge detection will also recognize the return of the same viewer with a reappearance of the same AutoPortrait.

Rachel Watson  16 February 1999