Unsupervised inference approach to facial attractiveness

Information-theoretical analysis of the neural code for decoupled face representation

Processing faces accurately and efficiently is a key capability of humans and other animals that engage in sophisticated social tasks. Recent studies reported a decoupled coding for faces in the primate inferotemporal cortex, with two separate neural populations coding for the geometric position of (texture-free) facial landmarks and for the image texture at fixed landmark positions, respectively. Here, we formally assess the efficiency of this decoupled coding by appealing to the information-theoretic notion of description length, which quantifies the amount of information that is saved when encoding novel facial images, with a given precision. We show that despite decoupled coding describes the facial images in terms of two sets of principal components (of landmark shape and image texture), it is more efficient (i.e., yields more information compression) than the encoding in terms of the image principal components only, which corresponds to the widely used eigenface method. The advantage of decoupled coding over eigenface coding increases with image resolution and is especially prominent when coding variants of training set images that only differ in facial expressions. Moreover, we demonstrate that decoupled coding entails better performance in three different tasks: the representation of facial images, the (daydream) sampling of novel facial images, and the recognition of facial identities and gender. In summary, our study provides a first principle perspective on the efficiency and accuracy of the decoupled coding of facial stimuli reported in the primate inferotemporal cortex.

A maximum entropy approach for the modelling of car-sharing parking dynamics

The science of cities is a relatively new and interdisciplinary topic aimed at studying and characterizing the collective processes that shape the growth and dynamics of urban populations. Amongst other open problems, the forecast of mobility trends in urban spaces is a lively research topic that aims at assisting the design and implementation of efficient transportation policies and inclusive urban planning. To this end, many Machine-Learning models have been put forward to predict mobility patterns. However, most of them are not interpretable -as they build on complex hidden representations of the system configurations- or do not allow for model inspection, thus limiting our understanding of the underlying mechanisms driving the citizen's daily routines. Here, we tackle this problem by building a fully interpretable statistical model that, incorporating only the minimum number of constraints, can predict different phenomena arising in the city. Using data on the movements of car-sharing vehicles in several Italian cities, we infer a model using the Maximum Entropy (MaxEnt) principle. The model allows for an accurate spatio-temporal prediction of car-sharing vehicles' presence in different city areas and, thanks to its simple yet general formulation, to precisely perform anomaly detection (e.g., detect strikes and bad weather conditions from car-sharing data only). We compare the forecasting capabilities of our model with different state-of-the-art models explicitly made for time-series forecasting: SARIMA models and Deep Learning Models. We find that MaxEnt models are highly predictive, outperforming SARIMAs while having similar performances of deep Neural Networks - but with advantages of being more interpretable, more flexibile-i.e., they can be applied to different tasks- and being computationally efficient. Our results show that statistical inference might play a fundamental role in building robust and general models describing urban systems phenomena.

Subjectivity and complexity of facial attractiveness

The origin and meaning of facial beauty represent a longstanding puzzle. Despite the profuse literature devoted to facial attractiveness, its very nature, its determinants and the nature of inter-person differences remain controversial issues. Here we tackle such questions proposing a novel experimental approach in which human subjects, instead of rating natural faces, are allowed to efficiently explore the face-space and "sculpt" their favorite variation of a reference facial image. The results reveal that different subjects prefer distinguishable regions of the face-space, highlighting the essential subjectivity of the phenomenon. The different sculpted facial vectors exhibit strong correlations among pairs of facial distances, characterising the underlying universality and complexity of the cognitive processes, and the relative relevance and robustness of the different facial distances.