W1: Face Recognition Flashcards
(22 cards)
What does the specificity account in facial recognition state?
States that neuro-cognitive mechanisms are selectively involved in processing faces per second and have little, if any role in processing non-face stimuli
What does the expertise account in facial recognition state?
States that neuro cognitive mechanisms involved in face processing are elicited for all prototype defined stimuli i.e. ones that have a shared configuration, as long as we have been PRE-EXPOSED to them
What is the face inversion effect and who proposed it?
Decrease in performance at recognising and discriminating between inverted faces compared to upright faces (Yin, 1969)
How does the effect of expertise hypothesis explain the face inversion effect?
We have lots of experience with upright faces and this builds up expertise for this class of stimuli resulting in better recognition of upright faces only
How does the specificity hypothesis explain the face inversion effect?
Argues that faces are ‘special’ i.e. having dedicated neural mechanisms specific to faces which only apply to upright faces
How did Yin (1969) test the specificity account of the FIE?
Black + white pictures (face of adult males w neutral expression, houses, airplanes)
Ppts shown individual pictures (study phase) and presented with pairs (recognition task). In the recognition task, they indicated the one pair they thought they had seen in the study phase. Orientation was both either upright or inverted.
Each ppt looked at 40 pics, rate of 3 secs per pic. Recognition task- 24 pairs, each pair containing 1 old picture from study phase and a new picture, asked which was old one. Four blocks of study phase + recognition task, two blocks upright and two inverted.
What did Yin (1969) find when testing the specificity account of the FIE?
They found a much higher amount of mean errors for recognition of inverted faces they had seen compared to upright faces, in comparison to upright/inverted houses or planes
How did Diamond and Carey (1986) test the expertise account of the face inversion effect?
Used photographs of dogs and human faces. Presented to novices and dog experts (of specific breeds presented).
Ppts instructed to remember each photo- inspected one series and immediately shown next series (humans or dogs depending on order). Two series (U or I) of 36 faces and two series of 18 dogs (U or I).
What did Diamond and Carey (1986) find when testing the expertise account of the face inversion effect?
Significant three way interaction of a subject group, material and orientation. Experts showed a similar inversion effect for faces and dogs whereas novels (not experts in dog breeds) did not. Suggests it is possible to obtain inversion effect for dog images similar to faces, ONLY if ppts are experts in the topic area.
What did Diamond and Carey argue we rely on to recognise faces?
We rely on our experience with the configural information
What are first and second order relations in configural processing?
First order relations- spatial relationships among the main features within a stimulus e.g. distance between eyes
Second order relations- variations in first order relations relative to the prototype for stimulus set
What areas of the face would require holistic and featural processing?
Holistic- entire face + shape
Featural- eyes and mouth
What is the difference between configural and featural/analytical processing?
Configural- Used to process faces- perception of spatial relationships amongst main features within a stimulus
Featural- perception of each specific feature in isolation from the rest, mainly used for objects
What is the Thatcher Illusion (Thompson, 2009)?
Phenomenon where it becomes more difficult to detect local feature changes in an upside down face compared to an upright face
What did Civile and Cooke et al’s (2020) experiment reveal about the Thatcher effect?
They found that the inversion effect for Thatcherised faces was significantly reduced than the inversion effect for regular faces
How did Parr and Heintz (2006) measure facial recognition and orientation in chimps?
Aimed to examine the influence of orientation on visual processing of faces in 6 chimpanzees. Five angles of orientation selected for two stimuli sets: unfamiliar chimpanzee faces and houses (as they represent a non expert category which shares several internal features/configurations with faces). Presented with no blurring, minimal blurring and severe blurring to suggest issues of configural processing. Samples were upright and comparisons inverted. Chimps had to move cursor on a joystick to contact the comparison image that matched the example- correct responses rewarded with food and incorrect five second ITI.
What did Parr and Heintz (2006) find in their study of facial recognition and orientation in chimps?
They found a face inversion effect for upright faces but not for houses, therefore supporting the FIE.
How did Parr et al. (2006) measure first vs. second order configuration in their second experiment of facial recognition in chimps?
Scrambling faces manipulation- therefore disrupting second order of configuration- took individual features of chimp face features and edited them on pictures to be moved around. Chimps had to first orient sample image, and then presented with experiment. Stimuli consisted of manipulated face image and unaltered face image in the corner- chimp had to contact comparison image that matched sample, given food if completed.
What did Parr et al. (2006) find when measuring first vs. second order configuration in their second experiment of facial recognition in chimps?
Performance was not v high overall for split and rearranged faces but higher for trials which presented inner face features- supporting the importance of second order relational features in chimpanzee face processing
What did Parr et al’s (2006) third experiment into facial recognition in chimps measure?
Digitized photographs of unfamiliar chimp faces. 40 unique trials- ten examples of four face categories presented. Ten were unaltered control trials where sample and correct comparison pair showed identical photos of different individuals. In 10 trials- same photos were edited so eyes of chimp in photo was masked with a black rectangle. In another 10- photo pixelated slightly, in another 10- pixilated a lot. Totals to 40 trials
What did Parr et al’s (2006) third experiment into facial recognition in chimps measure?
Lowest scores in large pixels for facial recognition which suggests first and second order relations are key in recognition. Difference between control image and blacking out eyes was not that much.
Summarise the findings of Parr et al.’s (2006) three studies
1- inversion effect suggests chimps, like humans, show facial configural processing
2- shows signif impairments when faces were manipulated to disrupt second order relations
3- revealed that pixelating faces using a large radius (affecting both first and second order relations) affects subject’s recognition performance
