In recent years, high-performance video recording devices have become ubiquitous, posing an unprecedented challenge to preserving personal privacy. As a result, privacy-preserving video systems have been receiving increased attention. In this paper, we present a novel privacy-preserving video algorithm that uses semantic segmentation to identify regions of interest, which are then anonymized with an adaptive blurring algorithm. This algorithm addresses two of the most important shortcomings of existing solutions: it is multi-scale, meaning it can identify and uniformly anonymize objects of different scales in the same image, and it is class-generic, so it can be used to anonymize any class of objects of interest. We show experimentally that our algorithm achieves excellent anonymity while preserving meaning in the visual data processed.

Automation and autonomous systems are quickly becoming a more engrained aspect of modern society. The need for effective, secure computer code in a timely manner has led to the creation of automated code repair techniques to resolve issues quickly. However, the research to date has largely ignored the human factors aspects of automated code repair. The current study explored trust perceptions, reuse intentions, and trust intentions in code repair with human generated patches versus automated code repair patches. In addition, comments in the headers were manipulated to determine the effect of the presence or absence of comments in the header of the code. Participants were 51 programmers with at least 3 years’ experience and knowledge of the C programming language. Results indicated only repair source (human vs. automated code repair) had a significant influence on trust perceptions and trust intentions. Specifically, participants consistently reported higher levels of perceived trustworthiness, intentions to reuse, and trust intentions for human referents compared to automated code repair. No significant effects were found for comments in the headers.

Open-source software (OSS) is a key aspect of software creation. However, little is known about programmers’ decisions to trust software from OSS websites. The current study emulated OSS websites and manipulated reputation and performance factors in the stimuli according to the heuristic-systematic processing model. We sampled professional programmers—with a minimum experience of three years—from Amazon Mechanical Turk (N = 38). We used a 3 × 3 within-subjects design to investigate the relationship between OSS reputation and performance on users’ time spent on code, the number of interface clicks, trustworthiness perceptions, and willingness to use OSS code. We found that participants spent more time on and clicked the interface more often for code that was high in reputation. Meta-information included with OSS tools was found to affect the degree to which computer programmers interact with and perceive online code repositories. Furthermore, participants reported higher levels of perceived trustworthiness in and trust toward highly reputable OSS code. Notably, we observed fewer significant main effects for the performance manipulation, which may correspond to participants considering performance attributes mainly within the context of reputation-relevant information. That is, the degree to which programmers investigate and then trust OSS code may depend on the initial reputation ratings.

Facial recognition tools are becoming exceptionally accurate in identifying people from images. However, this comes at the cost of privacy for users of online services with photo management (e.g. social media platforms). Particularly troubling is the ability to leverage unsupervised learning to recognize faces even when the user has not labeled their images. In this paper we propose Ulixes, a strategy to generate visually non-invasive facial noise masks that yield adversarial examples, preventing the formation of identifiable user clusters in the embedding space of facial encoders. This is applicable even when a user is unmasked and labeled images are available online. We demonstrate the effectiveness of Ulixes by showing that various classification and clustering methods cannot reliably label the adversarial examples we generate. We also study the effects of Ulixes in various black-box settings and compare it to the current state of the art in adversarial machine learning. Finally, we challenge the effectiveness of Ulixes against adversarially trained models and show that it is robust to countermeasures.

The exponential growth of data in the information age poses several threats to the privacy and safety of digital service users. Existing legislation, such as the GDPR in Europe and the CCPA in California, defines frameworks and guidelines to promote personal privacy but leaves freedom in the choice of means to achieve privacy. Data anonymization techniques remove information that can be used to identify individuals from the dataset, either through suppression, generalization, anatomization, permutation, or perturbation. Information suppression remains the most common, safe, and widely applicable anonymization method, though at a high data utility cost. In this paper, we argue that even information suppression may not be sufficient in some cases.We study the case of a dataset that describes the shopping habits of a grocery store’s customers. All identifiers and quasi-identifiers are removed from the dataset by suppression. However, by augmenting the data in a novel multi-step, iterative process, and building a neural network enriched with prior knowledge, we show that most sensitive information can be retrieved with an accuracy of 80%.