Face Detection In Pure Noise Images And Face Recognition Ability

Face is one specific stimulus. On one hand, most adults are exceptionally skilled at faceperception; on the other hand, face is different from other non-face stimuli in both neuralmechanism and cognitive processing manner. Considerable evidence from researches of facerecognition supports the hypothesis that humans have specialized cognitive and neuralmechanisms dedicated to the perception of faces (the face specificity hypothesis). Recent years,some researchers tested the face specificity hypothesis by investigating the specificity oftop-down face processing and made great progress. Is there any relation between facerecognition and top-down face processing? That means maybe there is some core cognitiveprocessing manner that makes the perceptual differences between face and other non-face stimuli.Discussion on this issue can help us know the nature of face processing mechanism. The presentresearch investigates whether individual differences in face detection in pure noise imagescorrelates with face recognition ability, and try to explain this correlation with holistic faceprocessing.The present research includes three parts. Experiment1tests the availability of detectionparadigm; experiment2investigates the relation between face detection and face recognitionability; experiment3and4investigate the relation between face detection and holistic faceprocessing, and with which we explain the relation between face detection and face recognitionability.The results are listed as below:(1) There is a significant negative correlation between face detection and face recognitionability. That means individuals who are better at detecting face in pure noise images are worse atrecognizing faces.(2) There is a left side bias in eye movement when participants detecting face in pure noiseimages.(3) Distribution of fixation is more concentrated in individuals who are better at detectingface in pure noise images, and more dispersive in individuals who are worse at detecting face in pure noise images.(4) There is a significant negative correlation between face detection and holistic faceprocessing. That means individuals who are better at detecting face in pure noise images areworse able to process faces holistically, and individuals who are worse at detecting face in purenoise images are better able to process faces holistically.(5) Individual difference in holistic face processing is one reason of the negative correlationbetween face detection and face recognition ability.