Albert Solé‐Ribalta

Cristina Bustos, Carles Civit, Brian Du et al.

This work presents a study on how to exploit the CLIP embedding space to perform Visual Sentiment Analysis. We experiment with two architectures built on top of the CLIP embedding space, which we denote by CLIP-E. We train the CLIP-E models with WEBEmo, the largest publicly available and manually labeled benchmark for Visual Sentiment Analysis, and perform two sets of experiments. First, we test on WEBEmo and compare the CLIP-E architectures with state-of-the-art (SOTA) models and with CLIP Zero-Shot. Second, we perform cross dataset evaluation, and test the CLIP-E architectures trained with WEBEmo on other Visual Sentiment Analysis benchmarks. Our results show that the CLIP-E approaches outperform SOTA models in WEBEmo fine grained categorization, and they also generalize better when tested on datasets that have not been seen during training. Interestingly, we observed that for the FI dataset, CLIP Zero-Shot produces better accuracies than SOTA models and CLIP-E trained on WEBEmo. These results motivate several questions that we discuss in this paper, such as how we should design new benchmarks and evaluate Visual Sentiment Analysis, and whether we should keep designing tailored Deep Learning models for Visual Sentiment Analysis or focus our efforts on better using the knowledge encoded in large vision-language models such as CLIP for this task.

María J. Palazzi, Jordi Cabot, Javier Luis Cánovas Izquierdo et al.

The development Open Source Software fundamentally depends on the participation and commitment of volunteer developers to progress. Several works have presented strategies to increase the on-boarding and engagement of new contributors, but little is known on how these diverse groups of developers self-organise to work together. To understand this, one must consider that, on one hand, platforms like GitHub provide a virtually unlimited development framework: any number of actors can potentially join to contribute in a decentralised, distributed, remote, and asynchronous manner. On the other, however, it seems reasonable that some sort of hierarchy and division of labour must be in place to meet human biological and cognitive limits, and also to achieve some level of efficiency. These latter features (hierarchy and division of labour) should translate into recognisable structural arrangements when projects are represented as developer-file bipartite networks. In this paper we analyse a set of popular open source projects from GitHub, placing the accent on three key properties: nestedness, modularity and in-block nestedness -which typify the emergence of heterogeneities among contributors, the emergence of subgroups of developers working on specific subgroups of files, and a mixture of the two previous, respectively. These analyses show that indeed projects evolve into internally organised blocks. Furthermore, the distribution of sizes of such blocks is bounded, connecting our results to the celebrated Dunbar number both in off- and on-line environments. Our analyses create a link between bio-cognitive constraints, group formation and online working environments, opening up a rich scenario for future research on (online) work team assembly.

Cristina Bustos, Daniel Rhoads, Agata Lapedriza et al.

Increased interaction between and among pedestrians and vehicles in the crowded urban environments of today gives rise to a negative side-effect: a growth in traffic accidents, with pedestrians being the most vulnerable elements. Recent work has shown that Convolutional Neural Networks are able to accurately predict accident rates exploiting Street View imagery along urban roads. The promising results point to the plausibility of aided design of safe urban landscapes, for both pedestrians and vehicles. In this paper, by considering historical accident data and Street View images, we detail how to automatically predict the impact (increase or decrease) of urban interventions on accident incidence. The results are positive, rendering an accuracies ranging from 60 to 80%. We additionally provide an interpretability analysis to unveil which specific categories of urban features impact accident rates positively or negatively. Considering the transportation network substrates (sidewalk and road networks) and their demand, we integrate these results to a complex network framework, to estimate the effective impact of urban change on the safety of pedestrians and vehicles. Results show that public authorities may leverage on machine learning tools to prioritize targeted interventions, since our analysis show that limited improvement is obtained with current tools. Further, our findings have a wider application range such as the design of safe urban routes for pedestrians or to the field of driver-assistance technologies.

Cristina Bustos, Daniel Rhoads, Albert Sole-Ribalta et al.

At the moment, urban mobility research and governmental initiatives are mostly focused on motor-related issues, e.g. the problems of congestion and pollution. And yet, we can not disregard the most vulnerable elements in the urban landscape: pedestrians, exposed to higher risks than other road users. Indeed, safe, accessible, and sustainable transport systems in cities are a core target of the UN's 2030 Agenda. Thus, there is an opportunity to apply advanced computational tools to the problem of traffic safety, in regards especially to pedestrians, who have been often overlooked in the past. This paper combines public data sources, large-scale street imagery and computer vision techniques to approach pedestrian and vehicle safety with an automated, relatively simple, and universally-applicable data-processing scheme. The steps involved in this pipeline include the adaptation and training of a Residual Convolutional Neural Network to determine a hazard index for each given urban scene, as well as an interpretability analysis based on image segmentation and class activation mapping on those same images. Combined, the outcome of this computational approach is a fine-grained map of hazard levels across a city, and an heuristic to identify interventions that might simultaneously improve pedestrian and vehicle safety. The proposed framework should be taken as a complement to the work of urban planners and public authorities.

Cristina Bustos, Neska Elhaouij, Albert Sole-Ribalta et al.

Several studies have shown the relevance of biosignals in driver stress recognition. In this work, we examine something important that has been less frequently explored: We develop methods to test if the visual driving scene can be used to estimate a drivers' subjective stress levels. For this purpose, we use the AffectiveROAD video recordings and their corresponding stress labels, a continuous human-driver-provided stress metric. We use the common class discretization for stress, dividing its continuous values into three classes: low, medium, and high. We design and evaluate three computer vision modeling approaches to classify the driver's stress levels: (1) object presence features, where features are computed using automatic scene segmentation; (2) end-to-end image classification; and (3) end-to-end video classification. All three approaches show promising results, suggesting that it is possible to approximate the drivers' subjective stress from the information found in the visual scene. We observe that the video classification, which processes the temporal information integrated with the visual information, obtains the highest accuracy of $0.72$, compared to a random baseline accuracy of $0.33$ when tested on a set of nine drivers.

Nicola Rebagliati, Albert Solé-Ribalta, Marcello Pelillo et al.

In structural pattern recognition, given a set of graphs, the computation of a Generalized Median Graph is a well known problem. Some methods approach the problem by assuming a relation between the Generalized Median Graph and the Common Labelling problem. However, this relation has still not been formally proved. In this paper, we analyse such relation between both problems. The main result proves that the cost of the common labelling upper-bounds the cost of the median with respect to the given set. In addition, we show that the two problems are equivalent in some cases.