Memory and Perception-based Facial Image Reconstruction

Perceptual comparisons induce lasting and generalizing changes to face memory reports

Humans are often tasked to remember new faces so that they can recognize the faces later in time. Previous studies found that memory reports for basic visual features (e.g., colors and shapes) are susceptible to systematic distortions as a result of comparison with new visual input, especially when the input is perceived as similar to the memory. The current study tested whether this similarity-induced memory bias (SIMB) would also occur with more complex face stimuli. The results showed that faces that are just perceptually encoded into visual working memory as well as retrieved from visual long-term memory are also susceptible to SIMB. Furthermore, once induced, SIMB persisted over time across cues through which the face memory was accessed for memory report. These results demonstrate the generalizability of SIMB to more complex and practically relevant stimuli, and thus, suggest potential real-world implications.

Your face looks the same as before, only prettier: The facial skin homogeneity effects on face change detection and facial attractiveness perception

Previous studies suggested that (1) facial attractiveness perception can be increased with facial skin homogeneity improving; and (2) human's facial change detection increases along with facial skin homogeneity increases. However, it's unknown whether a face can be perceived prettier than it did before while still being considered as physically the same. It is possible that these two kinds of cognitive-aesthetic processing may have separate mathematical functions in psychophysical studies. In other words, human's facial attractiveness differentiation may be more sensitive than facial change detection. In this current study, we explored the above questions. Using three types of psychophysical techniques to manipulate facial skin homogeneity, we measured how participants' sensitivity to facial skin homogeneity and attractiveness change. Results showed a linear function curve for facial physical change detection and a logarithmic function curve was drawn in the forced-choice technique, which was the most sensitive one, indicating that participants can judge a face prettier than before without being aware of it has physically changed. Besides, two linear function curves were shown in the same/different technique and a rating technique. Taken together, this current study revealed that facial attractiveness can be enhanced and discriminated by improving facial skin homogeneity, without being realized by people with conscious awareness that the face has been changed.

The Self in the Mind's Eye: Revealing How We Truly See Ourselves Through Reverse Correlation

Is there a way to visually depict the image people "see" of themselves in their minds' eyes? And if so, what can these mental images tell us about ourselves? We used a computational reverse-correlation technique to explore individuals' mental "self-portraits" of their faces and body shapes in an unbiased, data-driven way (total N = 116 adults). Self-portraits were similar to individuals' real faces but, importantly, also contained clues to each person's self-reported personality traits, which were reliably detected by external observers. Furthermore, people with higher social self-esteem produced more true-to-life self-portraits. Unlike face portraits, body portraits had negligible relationships with individuals' actual body shape, but as with faces, they were influenced by people's beliefs and emotions. We show how psychological beliefs and attitudes about oneself bias the perceptual representation of one's appearance and provide a unique window into the internal mental self-representation-findings that have important implications for mental health and visual culture.

The Face of Image Reconstruction: Progress, Pitfalls, Prospects

Recent research has demonstrated that neural and behavioral data acquired in response to viewing face images can be used to reconstruct the images themselves. However, the theoretical implications, promises, and challenges of this direction of research remain unclear. We evaluate the potential of this research for elucidating the visual representations underlying face recognition. Specifically, we outline complementary and converging accounts of the visual content, the representational structure, and the neural dynamics of face processing. We illustrate how this research addresses fundamental questions in the study of normal and impaired face recognition, and how image reconstruction provides a powerful framework for uncovering face representations, for unifying multiple types of empirical data, and for facilitating both theoretical and methodological progress.

Photogrammetry vs CT Scan: Evaluation of Accuracy of a Low-Cost Three-Dimensional Acquisition Method for Forensic Facial Approximation

Photogrammetry is a three-dimensional acquisition method potentially applicable to the forensic field. This possibility requires the verification of its accuracy. In this study, 3D volumes of skulls are generated to compare the photogrammetry versus the CT scan. In order to provide eligible material to the photogrammetric software, photographs were captured at a distance of 30 cm from the skull placed on a support 1 m in height and illuminated with diffused laboratory ceiling artificial light. A Nikon Coolpix P7100 camera was used. Photographs capture common elements with the previous and the next photograph so as to allow the photogrammetric software to recognize these common points between photographs and create a 3D puzzle. The Zephyr Lite (3DFlow©) software was employed to register the 3D volume. CT-based skulls are taken as a metric reference. The photogrammetry-based skulls are then enlarged according to the measurements of some landmarks or Zygion and Zygion, the distance between end of nasal and base of nasal pyramid for frontal projection, and minimum breadth of the mandibular ramus for the right lateral projection. The accuracy of the photogrammetry is compared to that of the CT scan by measuring the 3D volumes of the skulls studied. Specific landmarks are used as reference points for the measures in both frontal and lateral views. Bland-Altman graph shows homogeneity. The mean difference (1.28 mm) indicates that the measurements taken on the photogrammetry-based skull tend to slightly overestimate compared with the measurements taken on the CT-based skull.

Modelling face memory reveals task-generalizable representations

Current cognitive theories are cast in terms of information-processing mechanisms that use mental representations^1-4. For example, people use their mental representations to identify familiar faces under various conditions of pose, illumination and ageing, or to draw resemblance between family members. Yet, the actual information contents of these representations are rarely characterized, which hinders knowledge of the mechanisms that use them. Here, we modelled the three-dimensional representational contents of 4 faces that were familiar to 14 participants as work colleagues. The representational contents were created by reverse-correlating identity information generated on each trial with judgements of the face's similarity to the individual participant's memory of this face. In a second study, testing new participants, we demonstrated the validity of the modelled contents using everyday face tasks that generalize identity judgements to new viewpoints, age and sex. Our work highlights that such models of mental representations are critical to understanding generalization behaviour and its underlying information-processing mechanisms.