Bruno Lepri

Aditya Gulati, Bruno Lepri, Nuria Oliver

In recent years, there have been significant advancements in computer vision which have led to the widespread deployment of image recognition and generation systems in socially relevant applications, from hiring to security screening. However, the prevalence of biases within these systems has raised significant ethical and social concerns. The most extensively studied biases in this context are related to gender, race and age. Yet, other biases are equally pervasive and harmful, such as lookism, i.e., the preferential treatment of individuals based on their physical appearance. Lookism remains under-explored in computer vision but can have profound implications not only by perpetuating harmful societal stereotypes but also by undermining the fairness and inclusivity of AI technologies. Thus, this paper advocates for the systematic study of lookism as a critical bias in computer vision models. Through a comprehensive review of existing literature, we identify three areas of intersection between lookism and computer vision. We illustrate them by means of examples and a user study. We call for an interdisciplinary approach to address lookism, urging researchers, developers, and policymakers to prioritize the development of equitable computer vision systems that respect and reflect the diversity of human appearances.

Chenxi Li, Weijie Wang, Qiang Li et al.

Text-driven object insertion in 3D scenes is an emerging task that enables intuitive scene editing through natural language. However, existing 2D editing-based methods often rely on spatial priors such as 2D masks or 3D bounding boxes, and they struggle to ensure consistency of the inserted object. These limitations hinder flexibility and scalability in real-world applications. In this paper, we propose FreeInsert, a novel framework that leverages foundation models including MLLMs, LGMs, and diffusion models to disentangle object generation from spatial placement. This enables unsupervised and flexible object insertion in 3D scenes without spatial priors. FreeInsert starts with an MLLM-based parser that extracts structured semantics, including object types, spatial relationships, and attachment regions, from user instructions. These semantics guide both the reconstruction of the inserted object for 3D consistency and the learning of its degrees of freedom. We leverage the spatial reasoning capabilities of MLLMs to initialize object pose and scale. A hierarchical, spatially aware refinement stage further integrates spatial semantics and MLLM-inferred priors to enhance placement. Finally, the appearance of the object is improved using the inserted-object image to enhance visual fidelity. Experimental results demonstrate that FreeInsert achieves semantically coherent, spatially precise, and visually realistic 3D insertions without relying on spatial priors, offering a user-friendly and flexible editing experience.

Massimiliano Luca, Ciro Beneduce, Bruno Lepri et al.

With the wide and cross-domain adoption of Large Language Models, it becomes crucial to assess to which extent the statistical correlations in training data, which underlie their impressive performance, hide subtle and potentially troubling biases. Gender bias in LLMs has been widely investigated from the perspectives of works, hobbies, and emotions typically associated with a specific gender. In this study, we introduce a novel perspective. We investigate whether LLMs can predict an individual's gender based solely on online shopping histories and whether these predictions are influenced by gender biases and stereotypes. Using a dataset of historical online purchases from users in the United States, we evaluate the ability of six LLMs to classify gender and we then analyze their reasoning and products-gender co-occurrences. Results indicate that while models can infer gender with moderate accuracy, their decisions are often rooted in stereotypical associations between product categories and gender. Furthermore, explicit instructions to avoid bias reduce the certainty of model predictions, but do not eliminate stereotypical patterns. Our findings highlight the persistent nature of gender biases in LLMs and emphasize the need for robust bias-mitigation strategies.

Christos Perentis, Michele Vescovi, Chiara Leonardi et al.

The wide adoption of mobile devices and social media platforms have dramatically increased the collection and sharing of personal information. More and more frequently, users are called to take decisions concerning the disclosure of their personal information. In this study, we investigate the factors affecting users' choices toward the disclosure of their personal data, including not only their demographic and self-reported individual characteristics, but also their social interactions and their mobility patterns inferred from months of mobile phone data activity. We report the findings of a field-study conducted with a community of 63 subjects provided with (i) a smart-phone and (ii) a Personal Data Store (PDS) enabling them to control the disclosure of their data. We monitor the sharing behavior of our participants through the PDS, and evaluate the contribution of different factors affecting their disclosing choices of location and social interaction data. Our analysis shows that social interaction inferred by mobile phones is an important factor revealing willingness to share, regardless of the data type. In addition, we provide further insights on the individual traits relevant to the prediction of sharing behavior.

Andrey Bogomolov, Bruno Lepri, Fabio Pianesi

We present a new similarity measure based on information theoretic measures which is superior than Normalized Compression Distance for clustering problems and inherits the useful properties of conditional Kolmogorov complexity. We show that Normalized Compression Dictionary Size and Normalized Compression Dictionary Entropy are computationally more efficient, as the need to perform the compression itself is eliminated. Also they scale linearly with exponential vector size growth and are content independent. We show that normalized compression dictionary distance is compressor independent, if limited to lossless compressors, which gives space for optimizations and implementation speed improvement for real-time and big data applications. The introduced measure is applicable for machine learning tasks of parameter-free unsupervised clustering, supervised learning such as classification and regression, feature selection, and is applicable for big data problems with order of magnitude speed increase.