Unraveling Human Perception of Facial Aging Using Eye Gaze

Subjective report of eye fixations during serial search

Abstract: Humans readily introspect upon their thoughts and their behavior, but how reliable are these subjective reports? In the present study, we explored the consistencies of and differences between the observer’s subjective report and actual behavior within a single trial. On each trial of a serial search task, we recorded eye movements and the participants’ beliefs of where their eyes moved. The comparison of reported versus real eye movements revealed that subjects successfully reported a subset of their eye movements. Limits in subjective reports stemmed from both the number and the type of eye movements. Furthermore, subjects sometimes reported eye movements they actually never made. A detailed examination of these reports suggests that they could reflect covert shifts of attention during overt serial search. Our data provide quantitative and qualitative measures of observers’ subjective reports and reveal experimental effects of visual search that would otherwise be inaccessible.

On Recognizing Face Images With Weight and Age Variations

Abstract: With the increase in age, there are changes in skeletal structure, muscle mass, and body fat. For recognizing faces with age variations, researchers have generally focused on the skeletal structure and muscle mass. However, the effect of change in body fat has not been studied with respect to face recognition. In this paper, we incorporate weight information to improve the performance of face recognition with age variations. The proposed algorithm utilizes neural network and random decision forest to encode age variations across different weight categories. The results are reported on the WhoIsIt database prepared by the authors containing 1109 images from 110 individuals with age and weight variations. The comparison with existing state-of-the-art algorithms and commercial system on WhoIsIt and FG-Net databases shows that the proposed algorithm outperforms existing algorithms significantly.

Constructing models of user and task characteristics from eye gaze data for user-adaptive information highlighting

Abstract: A user-adaptive information visualization system capable of learning models of users and the visualization tasks they perform could provide interventions optimized for helping specific users in specific task contexts. In this paper, we investigate the accuracy of predicting visualization tasks, user performance on tasks, and user traits from gaze data. We show that predictions made with a logistic regression model are significantly better than a baseline classifier, with particularly strong results for predicting task type and user performance. Furthermore, we compare classifiers built with interface-independent and interface-dependent features, and show that the interface-independent features are comparable or superior to interface-dependent ones. Finally, we discuss how the accuracy of predictive models is affected if they are trained with data from trials that had highlighting interventions added to the visualization.

Facial Contrast Is a Cross-Cultural Cue for Perceiving Age.

Abstract: Age is a fundamental social dimension and a youthful appearance is of importance for many individuals, perhaps because it is a relevant predictor of aspects of health, facial attractiveness and general well-being. We recently showed that facial contrast – the color and luminance difference between facial features and the surrounding skin- is age-related and cue to age perception of Caucasian women. Specifically, aspects of facial contrast decrease with age in Caucasian women, and Caucasian female faces with higher contrast look younger (Porcheron et al, 2013). Here we investigated faces of other ethnic groups and raters of other cultures to see whether facial contrast is a cross-cultural youth related attribute. Using large sets of full face color photographs of Chinese, Latin American and black South African women, aged 20 to 80, we measured the luminance and color contrast between the facial features (the eyes, the lips and the brows) and the surrounding skin. Most aspects of facial contrast that were previously found to decrease with age in Caucasian women were also found to decrease with age in the other ethnic groups. Though the overall pattern of changes with age was common to all women, there were also some differences between the groups. In a separate study, individual faces of the 4 ethnic groups were perceived younger by French and Chinese participants when the aspects of facial contrast that vary with age in the majority of faces were artificially increased, but older when they were artificially decreased. Altogether these findings indicate that facial contrast is a cross-cultural cue to youthfulness. Because cosmetics were shown to enhance facial contrast, this work provides some support for the notion that a universal function of cosmetics is to make female faces look younger.

Multiple Projective Dictionary Learning to Detect Plastic Surgery for Face Verification

Abstract: Researchers have shown that the changes in face features due to plastic surgery can be modeled as a covariate that reduces the ability of algorithms to recognize a person’s identity. Traditional dictionary learning methods learn a sparse representation using $l_{0}$ and $l_{1}$ norms that are computationally expensive. This paper presents a multiple projective dictionary learning (MPDL) framework that does not require the computation of $l_{0}$ and $l_{1}$ norms. We propose a novel solution to discriminate plastic surgery faces from regular faces by learning representations of local and global plastic surgery faces using multiple projective dictionaries and by using compact binary face descriptors. Experimental results on the plastic surgery database show that the proposed MPDL framework is able to detect plastic surgery faces with a high accuracy of 97.96%. To verify the identity of a person, the detected plastic surgery faces are divided into local regions of interest (ROIs) that are likely to be altered by a particular plastic surgery. The cosine distance between the compact binary face descriptors is computed for each ROI in the detected plastic surgery faces. In addition, we compute the human visual system feature similarity score based on phase congruency and gradient magnitude between the same ROIs. The cosine distance scores and the feature similarity scores are combined to learn a support vector machine model to verify if the faces belong to the same person. We integrate our proposed MPDL framework for face verification with two commercial systems to demonstrate an improvement in verification performance on a combined database of plastic surgery and regular face images.