Giuseppe Serra

Alex Falcon, Swathikiran Sudhakaran, Giuseppe Serra et al.

Video retrieval using natural language queries has attracted increasing interest due to its relevance in real-world applications, from intelligent access in private media galleries to web-scale video search. Learning the cross-similarity of video and text in a joint embedding space is the dominant approach. To do so, a contrastive loss is usually employed because it organizes the embedding space by putting similar items close and dissimilar items far. This framework leads to competitive recall rates, as they solely focus on the rank of the groundtruth items. Yet, assessing the quality of the ranking list is of utmost importance when considering intelligent retrieval systems, since multiple items may share similar semantics, hence a high relevance. Moreover, the aforementioned framework uses a fixed margin to separate similar and dissimilar items, treating all non-groundtruth items as equally irrelevant. In this paper we propose to use a variable margin: we argue that varying the margin used during training based on how much relevant an item is to a given query, i.e. a relevance-based margin, easily improves the quality of the ranking lists measured through nDCG and mAP. We demonstrate the advantages of our technique using different models on EPIC-Kitchens-100 and YouCook2. We show that even if we carefully tuned the fixed margin, our technique (which does not have the margin as a hyper-parameter) would still achieve better performance. Finally, extensive ablation studies and qualitative analysis support the robustness of our approach. Code will be released at \url{https://github.com/aranciokov/RelevanceMargin-ICMR22}.

Alex Falcon, Giuseppe Serra, Oswald Lanz

Every hour, huge amounts of visual contents are posted on social media and user-generated content platforms. To find relevant videos by means of a natural language query, text-video retrieval methods have received increased attention over the past few years. Data augmentation techniques were introduced to increase the performance on unseen test examples by creating new training samples with the application of semantics-preserving techniques, such as color space or geometric transformations on images. Yet, these techniques are usually applied on raw data, leading to more resource-demanding solutions and also requiring the shareability of the raw data, which may not always be true, e.g. copyright issues with clips from movies or TV series. To address this shortcoming, we propose a multimodal data augmentation technique which works in the feature space and creates new videos and captions by mixing semantically similar samples. We experiment our solution on a large scale public dataset, EPIC-Kitchens-100, and achieve considerable improvements over a baseline method, improved state-of-the-art performance, while at the same time performing multiple ablation studies. We release code and pretrained models on Github at https://github.com/aranciokov/FSMMDA_VideoRetrieval.

Alex Falcon, Giuseppe Serra, Oswald Lanz

Due to the amount of videos and related captions uploaded every hour, deep learning-based solutions for cross-modal video retrieval are attracting more and more attention. A typical approach consists in learning a joint text-video embedding space, where the similarity of a video and its associated caption is maximized, whereas a lower similarity is enforced with all the other captions, called negatives. This approach assumes that only the video and caption pairs in the dataset are valid, but different captions - positives - may also describe its visual contents, hence some of them may be wrongly penalized. To address this shortcoming, we propose the Relevance-Aware Negatives and Positives mining (RANP) which, based on the semantics of the negatives, improves their selection while also increasing the similarity of other valid positives. We explore the influence of these techniques on two video-text datasets: EPIC-Kitchens-100 and MSR-VTT. By using the proposed techniques, we achieve considerable improvements in terms of nDCG and mAP, leading to state-of-the-art results, e.g. +5.3% nDCG and +3.0% mAP on EPIC-Kitchens-100. We share code and pretrained models at \url{https://github.com/aranciokov/ranp}.

Giuseppe Serra, Photios A. Stavrou, Marios Kountouris

In this paper, we study the computation of the rate-distortion-perception function (RDPF) for a multivariate Gaussian source under mean squared error (MSE) distortion and, respectively, Kullback-Leibler divergence, geometric Jensen-Shannon divergence, squared Hellinger distance, and squared Wasserstein-2 distance perception metrics. To this end, we first characterize the analytical bounds of the scalar Gaussian RDPF for the aforementioned divergence functions, also providing the RDPF-achieving forward "test-channel" realization. Focusing on the multivariate case, we establish that, for tensorizable distortion and perception metrics, the optimal solution resides on the vector space spanned by the eigenvector of the source covariance matrix. Consequently, the multivariate optimization problem can be expressed as a function of the scalar Gaussian RDPFs of the source marginals, constrained by global distortion and perception levels. Leveraging this characterization, we design an alternating minimization scheme based on the block nonlinear Gauss-Seidel method, which optimally solves the problem while identifying the Gaussian RDPF-achieving realization. Furthermore, the associated algorithmic embodiment is provided, as well as the convergence and the rate of convergence characterization. Lastly, for the "perfect realism" regime, the analytical solution for the multivariate Gaussian RDPF is obtained. We corroborate our results with numerical simulations and draw connections to existing results.

Massimiliano Incudini, Daniele Lizzio Bosco, Francesco Martini et al.

Quantum computing can empower machine learning models by enabling kernel machines to leverage quantum kernels for representing similarity measures between data. Quantum kernels are able to capture relationships in the data that are not efficiently computable on classical devices. However, there is no straightforward method to engineer the optimal quantum kernel for each specific use case. We present an approach to this problem, which employs optimization techniques, similar to those used in neural architecture search and AutoML, to automatically find an optimal kernel in a heuristic manner. To this purpose we define an algorithm for constructing a quantum circuit implementing the similarity measure as a combinatorial object, which is evaluated based on a cost function and then iteratively modified using a meta-heuristic optimization technique. The cost function can encode many criteria ensuring favorable statistical properties of the candidate solution, such as the rank of the Dynamical Lie Algebra. Importantly, our approach is independent of the optimization technique employed. The results obtained by testing our approach on a high-energy physics problem demonstrate that, in the best-case scenario, we can either match or improve testing accuracy with respect to the manual design approach, showing the potential of our technique to deliver superior results with reduced effort.

Giuseppe Serra, Mathias Niepert

Graph Neural Networks (GNNs) are a popular class of machine learning models. Inspired by the learning to explain (L2X) paradigm, we propose L2XGNN, a framework for explainable GNNs which provides faithful explanations by design. L2XGNN learns a mechanism for selecting explanatory subgraphs (motifs) which are exclusively used in the GNNs message-passing operations. L2XGNN is able to select, for each input graph, a subgraph with specific properties such as being sparse and connected. Imposing such constraints on the motifs often leads to more interpretable and effective explanations. Experiments on several datasets suggest that L2XGNN achieves the same classification accuracy as baseline methods using the entire input graph while ensuring that only the provided explanations are used to make predictions. Moreover, we show that L2XGNN is able to identify motifs responsible for the graph's properties it is intended to predict.

Alex Falcon, Giuseppe Serra, Sergio Escalera et al.

This report presents the technical details of our submission to the EPIC-Kitchens-100 Multi-Instance Retrieval Challenge 2022. To participate in the challenge, we designed an ensemble consisting of different models trained with two recently developed relevance-augmented versions of the widely used triplet loss. Our submission, visible on the public leaderboard, obtains an average score of 61.02% nDCG and 49.77% mAP.

Bhushan Kotnis, Kiril Gashteovski, Julia Gastinger et al.

With Human-Centric Research (HCR) we can steer research activities so that the research outcome is beneficial for human stakeholders, such as end users. But what exactly makes research human-centric? We address this question by providing a working definition and define how a research pipeline can be split into different stages in which human-centric components can be added. Additionally, we discuss existing NLP with HCR components and define a series of guiding questions, which can serve as starting points for researchers interested in exploring human-centric research approaches. We hope that this work would inspire researchers to refine the proposed definition and to pose other questions that might be meaningful for achieving HCR.

Simone Scaboro, Beatrice Portelli, Emmanuele Chersoni et al.

In the last decade, an increasing number of users have started reporting Adverse Drug Events (ADE) on social media platforms, blogs, and health forums. Given the large volume of reports, pharmacovigilance has focused on ways to use Natural Language Processing (NLP) techniques to rapidly examine these large collections of text, detecting mentions of drug-related adverse reactions to trigger medical investigations. However, despite the growing interest in the task and the advances in NLP, the robustness of these models in face of linguistic phenomena such as negations and speculations is an open research question. Negations and speculations are pervasive phenomena in natural language, and can severely hamper the ability of an automated system to discriminate between factual and nonfactual statements in text. In this paper we take into consideration four state-of-the-art systems for ADE detection on social media texts. We introduce SNAX, a benchmark to test their performance against samples containing negated and speculated ADEs, showing their fragility against these phenomena. We then introduce two possible strategies to increase the robustness of these models, showing that both of them bring significant increases in performance, lowering the number of spurious entities predicted by the models by 60% for negation and 80% for speculations.

Simone Scaboro, Beatrice Portellia, Emmanuele Chersoni et al.

Adverse Event (ADE) extraction is one of the core tasks in digital pharmacovigilance, especially when applied to informal texts. This task has been addressed by the Natural Language Processing community using large pre-trained language models, such as BERT. Despite the great number of Transformer-based architectures used in the literature, it is unclear which of them has better performances and why. Therefore, in this paper we perform an extensive evaluation and analysis of 19 Transformer-based models for ADE extraction on informal texts. We compare the performance of all the considered models on two datasets with increasing levels of informality (forums posts and tweets). We also combine the purely Transformer-based models with two commonly-used additional processing layers (CRF and LSTM), and analyze their effect on the models performance. Furthermore, we use a well-established feature importance technique (SHAP) to correlate the performance of the models with a set of features that describe them: model category (AutoEncoding, AutoRegressive, Text-to-Text), pretraining domain, training from scratch, and model size in number of parameters. At the end of our analyses, we identify a list of take-home messages that can be derived from the experimental data.

Beatrice Portelli, Simone Scaboro, Enrico Santus et al.

Medical term normalization consists in mapping a piece of text to a large number of output classes. Given the small size of the annotated datasets and the extremely long tail distribution of the concepts, it is of utmost importance to develop models that are capable to generalize to scarce or unseen concepts. An important attribute of most target ontologies is their hierarchical structure. In this paper we introduce a simple and effective learning strategy that leverages such information to enhance the generalizability of both discriminative and generative models. The evaluation shows that the proposed strategy produces state-of-the-art performance on seen concepts and consistent improvements on unseen ones, allowing also for efficient zero-shot knowledge transfer across text typologies and datasets.

Beatrice Portelli, Simone Scaboro, Emmanuele Chersoni et al.

This paper describes the models developed by the AILAB-Udine team for the SMM4H 22 Shared Task. We explored the limits of Transformer based models on text classification, entity extraction and entity normalization, tackling Tasks 1, 2, 5, 6 and 10. The main take-aways we got from participating in different tasks are: the overwhelming positive effects of combining different architectures when using ensemble learning, and the great potential of generative models for term normalization.

Zhao Xu, Daniel Onoro Rubio, Giuseppe Serra et al.

Deep latent generative models have attracted increasing attention due to the capacity of combining the strengths of deep learning and probabilistic models in an elegant way. The data representations learned with the models are often continuous and dense. However in many applications, sparse representations are expected, such as learning sparse high dimensional embedding of data in an unsupervised setting, and learning multi-labels from thousands of candidate tags in a supervised setting. In some scenarios, there could be further restriction on degree of sparsity: the number of non-zero features of a representation cannot be larger than a pre-defined threshold $L_0$. In this paper we propose a sparse deep latent generative model SDLGM to explicitly model degree of sparsity and thus enable to learn the sparse structure of the data with the quantified sparsity constraint. The resulting sparsity of a representation is not fixed, but fits to the observation itself under the pre-defined restriction. In particular, we introduce to each observation $i$ an auxiliary random variable $L_i$, which models the sparsity of its representation. The sparse representations are then generated with a two-step sampling process via two Gumbel-Softmax distributions. For inference and learning, we develop an amortized variational method based on MC gradient estimator. The resulting sparse representations are differentiable with backpropagation. The experimental evaluation on multiple datasets for unsupervised and supervised learning problems shows the benefits of the proposed method.

Daniele Lizzio Bosco, Lukasz Cincio, Giuseppe Serra et al.

Compiling quantum circuits into Clifford+$T$ gates is a central task for fault-tolerant quantum computing using stabilizer codes. In the near term, $T$ gates will dominate the cost of fault tolerant implementations, and any reduction in the number of such expensive gates could mean the difference between being able to run a circuit or not. While exact synthesis is exponentially hard in the number of qubits, local synthesis approaches are commonly used to compile large circuits by decomposing them into substructures. However, composing local methods leads to suboptimal compilations in key metrics such as $T$-count or circuit depth, and their performance strongly depends on circuit representation. In this work, we address this challenge by proposing \textsc{Q-PreSyn}, a strategy that, given a set of local edits preserving circuit equivalence, uses a RL agent to identify effective sequences of such actions and thereby obtain circuit representations that yield a reduced $T$-count upon synthesis. Experimental results of our proposed strategy, applied on top of well-known synthesis algorithms, show up to a $20\%$ reduction in $T$-count on circuits with up to 25 qubits, without introducing any additional approximation error prior to synthesis.

Giuseppe Serra, Photios A. Stavrou, Marios Kountouris

We study the computation of the rate-distortion-perception function (RDPF) for discrete memoryless sources subject to a single-letter average distortion constraint and a perception constraint belonging to the family of $f$-divergences. In this setting, the RDPF forms a convex programming problem for which we characterize optimal parametric solutions. We employ the developed solutions in an alternating minimization scheme, namely Optimal Alternating Minimization (OAM), for which we provide convergence guarantees. Nevertheless, the OAM scheme does not lead to a direct implementation of a generalized Blahut-Arimoto (BA) type of algorithm due to implicit equations in the iteration's structure. To overcome this difficulty, we propose two alternative minimization approaches whose applicability depends on the smoothness of the used perception metric: a Newton-based Alternating Minimization (NAM) scheme, relying on Newton's root-finding method for the approximation of the optimal solution of the iteration, and a Relaxed Alternating Minimization (RAM) scheme, based on relaxing the OAM iterates. We show, by deriving necessary and sufficient conditions, that both schemes guarantee convergence to a globally optimal solution. We also provide sufficient conditions on the distortion and perception constraints, which guarantee that the proposed algorithms converge exponentially fast in the number of iteration steps. We corroborate our theoretical results with numerical simulations and establish connections with existing results.

Ali Abdari, Alex Falcon, Giuseppe Serra

Nowadays, many people frequently have to search for new accommodation options. Searching for a suitable apartment is a time-consuming process, especially because visiting them is often mandatory to assess the truthfulness of the advertisements found on the Web. While this process could be alleviated by visiting the apartments in the metaverse, the Web-based recommendation platforms are not suitable for the task. To address this shortcoming, in this paper, we define a new problem called text-to-apartment recommendation, which requires ranking the apartments based on their relevance to a textual query expressing the user's interests. To tackle this problem, we introduce FArMARe, a multi-task approach that supports cross-modal contrastive training with a furniture-aware objective. Since public datasets related to indoor scenes do not contain detailed descriptions of the furniture, we collect and annotate a dataset comprising more than 6000 apartments. A thorough experimentation with three different methods and two raw feature extraction procedures reveals the effectiveness of FArMARe in dealing with the problem at hand.

Jacopo Cossio, Daniele Lizzio Bosco, Riccardo Romanello et al.

Quantum circuit optimization is a central task in Quantum Computing, as current Noisy Intermediate Scale Quantum devices suffer from error propagation that often scales with the number of operations. Among quantum operations, the CNOT gate is of fundamental importance, being the only 2-qubit gate in the universal Clifford+T set. The problem of CNOT gates minimization has been addressed by heuristic algorithms such as the well-known Patel-Markov-Hayes (PMH) for linear reversible synthesis (i.e., CNOT minimization with no topological constraints), and more recently by Reinforcement Learning (RL) based strategies in the more complex case of topology-aware synthesis, where each CNOT can act on a subset of all qubits pairs. In this work we introduce AlphaCNOT, a RL framework based on Monte Carlo Tree Search (MCTS) that address effectively the CNOT minimization problem by modeling it as a planning problem. In contrast to other RL- based solution, our method is model-based, i.e. it can leverage lookahead search to evaluate future trajectories, thus finding more efficient sequences of CNOTs. Our method achieves a reduction of up to 32% in CNOT gate count compared to PMH baseline on linear reversible synthesis, while in the constraint version we report a consistent gate count reduction on a variety of topologies with up to 8 qubits, with respect to state-of-the-art RL-based solutions. Our results suggest the combination of RL with search-based strategies can be applied to different circuit optimization tasks, such as Clifford minimization, thus fostering the transition toward the "quantum utility" era.

Giuseppe Serra, Ben Werner, Florian Buettner

Many real-world applications require machine-learning models to be able to deal with non-stationary data distributions and thus learn autonomously over an extended period of time, often in an online setting. One of the main challenges in this scenario is the so-called catastrophic forgetting (CF) for which the learning model tends to focus on the most recent tasks while experiencing predictive degradation on older ones. In the online setting, the most effective solutions employ a fixed-size memory buffer to store old samples used for replay when training on new tasks. Many approaches have been presented to tackle this problem. However, it is not clear how predictive uncertainty information for memory management can be leveraged in the most effective manner and conflicting strategies are proposed to populate the memory. Are the easiest-to-forget or the easiest-to-remember samples more effective in combating CF? Starting from the intuition that predictive uncertainty provides an idea of the samples' location in the decision space, this work presents an in-depth analysis of different uncertainty estimates and strategies for populating the memory. The investigation provides a better understanding of the characteristics data points should have for alleviating CF. Then, we propose an alternative method for estimating predictive uncertainty via the generalised variance induced by the negative log-likelihood. Finally, we demonstrate that the use of predictive uncertainty measures helps in reducing CF in different settings.

Alex Falcon, Giuseppe Serra

In this report, we present the technical details of our submission to the EPIC-Kitchens-100 Multi-Instance Retrieval Challenge 2023. To participate in the challenge, we ensembled two models trained with two different loss functions on 25% of the training data. Our submission, visible on the public leaderboard, obtains an average score of 56.81% nDCG and 42.63% mAP.

Ali Abdari, Alex Falcon, Giuseppe Serra

Every day, a large amount of educational content is uploaded online across different areas, including agriculture and gardening. When these videos or materials are grouped meaningfully, they can make learning easier and more effective. One promising way to organize and enrich such content is through the Metaverse, which allows users to explore educational experiences in an interactive and immersive environment. However, searching for relevant Metaverse scenarios and finding those matching users' interests remains a challenging task. A first step in this direction has been done recently, but existing datasets are small and not sufficient for training advanced models. In this work, we make two main contributions: first, we introduce a new dataset containing 457 agricultural-themed virtual museums (AgriMuseums), each enriched with textual descriptions; and second, we propose a hierarchical vision-language model to represent and retrieve relevant AgriMuseums using natural language queries. In our experimental setting, the proposed method achieves up to about 62\% R@1 and 78\% MRR, confirming its effectiveness, and it also leads to improvements on existing benchmarks by up to 6\% R@1 and 11\% MRR. Moreover, an extensive evaluation validates our design choices. Code and dataset are available at https://github.com/aliabdari/Agricultural_Metaverse_Retrieval .

Daniele Lizzio Bosco, Shuteng Wang, Giuseppe Serra et al.

Recently, Quantum Visual Fields (QVFs) have shown promising improvements in model compactness and convergence speed for learning the provided 2D or 3D signals. Meanwhile, novel-view synthesis has seen major advances with Neural Radiance Fields (NeRFs), where models learn a compact representation from 2D images to render 3D scenes, albeit at the cost of larger models and intensive training. In this work, we extend the approach of QVFs by introducing QNeRF, the first hybrid quantum-classical model designed for novel-view synthesis from 2D images. QNeRF leverages parameterised quantum circuits to encode spatial and view-dependent information via quantum superposition and entanglement, resulting in more compact models compared to the classical counterpart. We present two architectural variants. Full QNeRF maximally exploits all quantum amplitudes to enhance representational capabilities. In contrast, Dual-Branch QNeRF introduces a task-informed inductive bias by branching spatial and view-dependent quantum state preparations, drastically reducing the complexity of this operation and ensuring scalability and potential hardware compatibility. Our experiments demonstrate that -- when trained on images of moderate resolution -- QNeRF matches or outperforms classical NeRF baselines while using less than half the number of parameters. These results suggest that quantum machine learning can serve as a competitive alternative for continuous signal representation in mid-level tasks in computer vision, such as 3D representation learning from 2D observations.

Ali Abdari, Alex Falcon, Giuseppe Serra

Recently, the Metaverse is becoming increasingly attractive, with millions of users accessing the many available virtual worlds. However, how do users find the one Metaverse which best fits their current interests? So far, the search process is mostly done by word of mouth, or by advertisement on technology-oriented websites. However, the lack of search engines similar to those available for other multimedia formats (e.g., YouTube for videos) is showing its limitations, since it is often cumbersome to find a Metaverse based on some specific interests using the available methods, while also making it difficult to discover user-created ones which lack strong advertisement. To address this limitation, we propose to use language to naturally describe the desired contents of the Metaverse a user wishes to find. Second, we highlight that, differently from more conventional 3D scenes, Metaverse scenarios represent a more complex data format since they often contain one or more types of multimedia which influence the relevance of the scenario itself to a user query. Therefore, in this work, we create a novel task, called Text-to-Metaverse retrieval, which aims at modeling these aspects while also taking the cross-modal relations with the textual data into account. Since we are the first ones to tackle this problem, we also collect a dataset of 33000 Metaverses, each of which consists of a 3D scene enriched with multimedia content. Finally, we design and implement a deep learning framework based on contrastive learning, resulting in a thorough experimental setup.

Nour Jamoussi, Giuseppe Serra, Photios A. Stavrou et al.

Federated learning (FL) is a widely used and impactful distributed optimization framework that achieves consensus through averaging locally trained models. While effective, this approach may not align well with Bayesian inference, where the model space has the structure of a distribution space. Taking an information-geometric perspective, we reinterpret FL aggregation as the problem of finding the barycenter of local posteriors using a prespecified divergence metric, minimizing the average discrepancy across clients. This perspective provides a unifying framework that generalizes many existing methods and offers crisp insights into their theoretical underpinnings. We then propose BA-BFL, an algorithm that retains the convergence properties of Federated Averaging in non-convex settings. In non-independent and identically distributed scenarios, we conduct extensive comparisons with statistical aggregation techniques, showing that BA-BFL achieves performance comparable to state-of-the-art methods while offering a geometric interpretation of the aggregation phase. Additionally, we extend our analysis to Hybrid Bayesian Deep Learning, exploring the impact of Bayesian layers on uncertainty quantification and model calibration.

Riccardo Romanello, Daniele Lizzio Bosco, Jacopo Cossio et al.

CNOT gates are fundamental to quantum computing, as they facilitate entanglement, a crucial resource for quantum algorithms. Certain classes of quantum circuits are constructed exclusively from CNOT gates. Given their widespread use, it is imperative to minimise the number of CNOT gates employed. This problem, known as CNOT minimisation, remains an open challenge, with its computational complexity yet to be fully characterised. In this work, we introduce a novel reinforcement learning approach to address this task. Instead of training multiple reinforcement learning agents for different circuit sizes, we use a single agent up to a fixed size $m$. Matrices of sizes different from m are preprocessed using either embedding or Gaussian striping. To assess the efficacy of our approach, we trained an agent with m = 8, and evaluated it on matrices of size n that range from 3 to 15. The results we obtained show that our method overperforms the state-of-the-art algorithm as the value of n increases.

Giuseppe Serra, Florian Buettner

Continual Learning (CL) is recently gaining increasing attention for its ability to enable a single model to learn incrementally from a sequence of new classes. In this scenario, it is important to keep consistent predictive performance across all the classes and prevent the so-called Catastrophic Forgetting (CF). However, in safety-critical applications, predictive performance alone is insufficient. Predictive models should also be able to reliably communicate their uncertainty in a calibrated manner - that is, with confidence scores aligned to the true frequencies of target events. Existing approaches in CL address calibration primarily from a data-centric perspective, relying on a single temperature shared across all tasks. Such solutions overlook task-specific differences, leading to large fluctuations in calibration error across tasks. For this reason, we argue that a more principled approach should adapt the temperature according to the distance to the current task. However, the unavailability of the task information at test time/during deployment poses a major challenge to achieve the intended objective. For this, we propose Distance-Aware Temperature Scaling (DATS), which combines prototype-based distance estimation with distance-aware calibration to infer task proximity and assign adaptive temperatures without prior task information. Through extensive empirical evaluation on both standard benchmarks and real-world, imbalanced datasets taken from the biomedical domain, our approach demonstrates to be stable, reliable and consistent in reducing calibration error across tasks compared to state-of-the-art approaches.

Nour Jamoussi, Giuseppe Serra, Photios A. Stavrou et al.

Bayesian Federated Learning (BFL) combines uncertainty modeling with decentralized training, enabling the development of personalized and reliable models under data heterogeneity and privacy constraints. Existing approaches typically rely on Markov Chain Monte Carlo (MCMC) sampling or variational inference, often incorporating personalization mechanisms to better adapt to local data distributions. In this work, we propose an information-geometric projection framework for personalization in parametric BFL. By projecting the global model onto a neighborhood of the user's local model, our method enables a tunable trade-off between global generalization and local specialization. Under mild assumptions, we show that this projection step is equivalent to computing a barycenter on the statistical manifold, allowing us to derive closed-form solutions and achieve cost-free personalization. We apply the proposed approach to a variational learning setup using the Improved Variational Online Newton (IVON) optimizer and extend its application to general aggregation schemes in BFL. Empirical evaluations under heterogeneous data distributions confirm that our method effectively balances global and local performance with minimal computational overhead.

Giuseppe Serra, Peter Tino, Zhao Xu et al.

Nowadays, neural network (NN) and deep learning (DL) techniques are widely adopted in many applications, including recommender systems. Given the sparse and stochastic nature of collaborative filtering (CF) data, recent works have critically analyzed the effective improvement of neural-based approaches compared to simpler and often transparent algorithms for recommendation. Previous results showed that NN and DL models can be outperformed by traditional algorithms in many tasks. Moreover, given the largely black-box nature of neural-based methods, interpretable results are not naturally obtained. Following on this debate, we first present a transparent probabilistic model that topologically organizes user and product latent classes based on the review information. In contrast to popular neural techniques for representation learning, we readily obtain a statistical, visualization-friendly tool that can be easily inspected to understand user and product characteristics from a textual-based perspective. Then, given the limitations of common embedding techniques, we investigate the possibility of using the estimated interpretable quantities as model input for a rating prediction task. To contribute to the recent debates, we evaluate our results in terms of both capacity for interpretability and predictive performances in comparison with popular text-based neural approaches. The results demonstrate that the proposed latent class representations can yield competitive predictive performances, compared to popular, but difficult-to-interpret approaches.

Simone Scaboro, Beatrice Portelli, Emmanuele Chersoni et al.

Adverse Drug Event (ADE) extraction models can rapidly examine large collections of social media texts, detecting mentions of drug-related adverse reactions and trigger medical investigations. However, despite the recent advances in NLP, it is currently unknown if such models are robust in face of negation, which is pervasive across language varieties. In this paper we evaluate three state-of-the-art systems, showing their fragility against negation, and then we introduce two possible strategies to increase the robustness of these models: a pipeline approach, relying on a specific component for negation detection; an augmentation of an ADE extraction dataset to artificially create negated samples and further train the models. We show that both strategies bring significant increases in performance, lowering the number of spurious entities predicted by the models. Our dataset and code will be publicly released to encourage research on the topic.

Kevin Roitero, Michael Soprano, Beatrice Portelli et al.

Recently, the misinformation problem has been addressed with a crowdsourcing-based approach: to assess the truthfulness of a statement, instead of relying on a few experts, a crowd of non-expert is exploited. We study whether crowdsourcing is an effective and reliable method to assess truthfulness during a pandemic, targeting statements related to COVID-19, thus addressing (mis)information that is both related to a sensitive and personal issue and very recent as compared to when the judgment is done. In our experiments, crowd workers are asked to assess the truthfulness of statements, and to provide evidence for the assessments. Besides showing that the crowd is able to accurately judge the truthfulness of the statements, we report results on workers behavior, agreement among workers, effect of aggregation functions, of scales transformations, and of workers background and bias. We perform a longitudinal study by re-launching the task multiple times with both novice and experienced workers, deriving important insights on how the behavior and quality change over time. Our results show that: workers are able to detect and objectively categorize online (mis)information related to COVID-19; both crowdsourced and expert judgments can be transformed and aggregated to improve quality; worker background and other signals (e.g., source of information, behavior) impact the quality of the data. The longitudinal study demonstrates that the time-span has a major effect on the quality of the judgments, for both novice and experienced workers. Finally, we provide an extensive failure analysis of the statements misjudged by the crowd-workers.

Beatrice Portelli, Daniele Passabì, Edoardo Lenzi et al.

In recent years, Internet users are reporting Adverse Drug Events (ADE) on social media, blogs and health forums. Because of the large volume of reports, pharmacovigilance is seeking to resort to NLP to monitor these outlets. We propose for the first time the use of the SpanBERT architecture for the task of ADE extraction: this new version of the popular BERT transformer showed improved capabilities with multi-token text spans. We validate our hypothesis with experiments on two datasets (SMM4H and CADEC) with different text typologies (tweets and blog posts), finding that SpanBERT combined with a CRF outperforms all the competitors on both of them.

Alex Falcon, Oswald Lanz, Giuseppe Serra

Video Question Answering (VideoQA) is a task that requires a model to analyze and understand both the visual content given by the input video and the textual part given by the question, and the interaction between them in order to produce a meaningful answer. In our work we focus on the Egocentric VideoQA task, which exploits first-person videos, because of the importance of such task which can have impact on many different fields, such as those pertaining the social assistance and the industrial training. Recently, an Egocentric VideoQA dataset, called EgoVQA, has been released. Given its small size, models tend to overfit quickly. To alleviate this problem, we propose several augmentation techniques which give us a +5.5% improvement on the final accuracy over the considered baseline.

Kevin Roitero, Michael Soprano, Beatrice Portelli et al.

Misinformation is an ever increasing problem that is difficult to solve for the research community and has a negative impact on the society at large. Very recently, the problem has been addressed with a crowdsourcing-based approach to scale up labeling efforts: to assess the truthfulness of a statement, instead of relying on a few experts, a crowd of (non-expert) judges is exploited. We follow the same approach to study whether crowdsourcing is an effective and reliable method to assess statements truthfulness during a pandemic. We specifically target statements related to the COVID-19 health emergency, that is still ongoing at the time of the study and has arguably caused an increase of the amount of misinformation that is spreading online (a phenomenon for which the term "infodemic" has been used). By doing so, we are able to address (mis)information that is both related to a sensitive and personal issue like health and very recent as compared to when the judgment is done: two issues that have not been analyzed in related work. In our experiment, crowd workers are asked to assess the truthfulness of statements, as well as to provide evidence for the assessments as a URL and a text justification. Besides showing that the crowd is able to accurately judge the truthfulness of the statements, we also report results on many different aspects, including: agreement among workers, the effect of different aggregation functions, of scales transformations, and of workers background / bias. We also analyze workers behavior, in terms of queries submitted, URLs found / selected, text justifications, and other behavioral data like clicks and mouse actions collected by means of an ad hoc logger.

Marco Menardi, Alex Falcon, Saida S. Mohamed et al.

Text to Image Synthesis refers to the process of automatic generation of a photo-realistic image starting from a given text and is revolutionizing many real-world applications. In order to perform such process it is necessary to exploit datasets containing captioned images, meaning that each image is associated with one (or more) captions describing it. Despite the abundance of uncaptioned images datasets, the number of captioned datasets is limited. To address this issue, in this paper we propose an approach capable of generating images starting from a given text using conditional GANs trained on uncaptioned images dataset. In particular, uncaptioned images are fed to an Image Captioning Module to generate the descriptions. Then, the GAN Module is trained on both the input image and the machine-generated caption. To evaluate the results, the performance of our solution is compared with the results obtained by the unconditional GAN. For the experiments, we chose to use the uncaptioned dataset LSUN bedroom. The results obtained in our study are preliminary but still promising.

Marco Zamprogno, Marco Passon, Niki Martinel et al.

In this paper we consider the problem of video-based person re-identification, which is the task of associating videos of the same person captured by different and non-overlapping cameras. We propose a Siamese framework in which video frames of the person to re-identify and of the candidate one are processed by two identical networks which produce a similarity score. We introduce an attention mechanisms to capture the relevant information both at frame level (spatial information) and at video level (temporal information given by the importance of a specific frame within the sequence). One of the novelties of our approach is given by a joint concurrent processing of both frame and video levels, providing in such a way a very simple architecture. Despite this fact, our approach achieves better performance than the state-of-the-art on the challenging iLIDS-VID dataset.

Marco Passon, Marco Lippi, Giuseppe Serra et al.

Internet users generate content at unprecedented rates. Building intelligent systems capable of discriminating useful content within this ocean of information is thus becoming a urgent need. In this paper, we aim to predict the usefulness of Amazon reviews, and to do this we exploit features coming from an off-the-shelf argumentation mining system. We argue that the usefulness of a review, in fact, is strictly related to its argumentative content, whereas the use of an already trained system avoids the costly need of relabeling a novel dataset. Results obtained on a large publicly available corpus support this hypothesis.

Marcella Cornia, Lorenzo Baraldi, Giuseppe Serra et al.

Image captioning has been recently gaining a lot of attention thanks to the impressive achievements shown by deep captioning architectures, which combine Convolutional Neural Networks to extract image representations, and Recurrent Neural Networks to generate the corresponding captions. At the same time, a significant research effort has been dedicated to the development of saliency prediction models, which can predict human eye fixations. Even though saliency information could be useful to condition an image captioning architecture, by providing an indication of what is salient and what is not, research is still struggling to incorporate these two techniques. In this work, we propose an image captioning approach in which a generative recurrent neural network can focus on different parts of the input image during the generation of the caption, by exploiting the conditioning given by a saliency prediction model on which parts of the image are salient and which are contextual. We show, through extensive quantitative and qualitative experiments on large scale datasets, that our model achieves superior performances with respect to captioning baselines with and without saliency, and to different state of the art approaches combining saliency and captioning.

Marcella Cornia, Lorenzo Baraldi, Giuseppe Serra et al.

Data-driven saliency has recently gained a lot of attention thanks to the use of Convolutional Neural Networks for predicting gaze fixations. In this paper we go beyond standard approaches to saliency prediction, in which gaze maps are computed with a feed-forward network, and present a novel model which can predict accurate saliency maps by incorporating neural attentive mechanisms. The core of our solution is a Convolutional LSTM that focuses on the most salient regions of the input image to iteratively refine the predicted saliency map. Additionally, to tackle the center bias typical of human eye fixations, our model can learn a set of prior maps generated with Gaussian functions. We show, through an extensive evaluation, that the proposed architecture outperforms the current state of the art on public saliency prediction datasets. We further study the contribution of each key component to demonstrate their robustness on different scenarios.

Marcella Cornia, Lorenzo Baraldi, Giuseppe Serra et al.

This paper presents a novel deep architecture for saliency prediction. Current state of the art models for saliency prediction employ Fully Convolutional networks that perform a non-linear combination of features extracted from the last convolutional layer to predict saliency maps. We propose an architecture which, instead, combines features extracted at different levels of a Convolutional Neural Network (CNN). Our model is composed of three main blocks: a feature extraction CNN, a feature encoding network, that weights low and high level feature maps, and a prior learning network. We compare our solution with state of the art saliency models on two public benchmarks datasets. Results show that our model outperforms under all evaluation metrics on the SALICON dataset, which is currently the largest public dataset for saliency prediction, and achieves competitive results on the MIT300 benchmark.

Stefano Alletto, Giuseppe Serra, Rita Cucchiara

With the spread of wearable devices and head mounted cameras, a wide range of application requiring precise user localization is now possible. In this paper we propose to treat the problem of obtaining the user position with respect to a known environment as a video registration problem. Video registration, i.e. the task of aligning an input video sequence to a pre-built 3D model, relies on a matching process of local keypoints extracted on the query sequence to a 3D point cloud. The overall registration performance is strictly tied to the actual quality of this 2D-3D matching, and can degrade if environmental conditions such as steep changes in lighting like the ones between day and night occur. To effectively register an egocentric video sequence under these conditions, we propose to tackle the source of the problem: the matching process. To overcome the shortcomings of standard matching techniques, we introduce a novel embedding space that allows us to obtain robust matches by jointly taking into account local descriptors, their spatial arrangement and their temporal robustness. The proposal is evaluated using unconstrained egocentric video sequences both in terms of matching quality and resulting registration performance using different 3D models of historical landmarks. The results show that the proposed method can outperform state of the art registration algorithms, in particular when dealing with the challenges of night and day sequences.

Lamberto Ballan, Marco Bertini, Giuseppe Serra et al.

Tagging of visual content is becoming more and more widespread as web-based services and social networks have popularized tagging functionalities among their users. These user-generated tags are used to ease browsing and exploration of media collections, e.g. using tag clouds, or to retrieve multimedia content. However, not all media are equally tagged by users. Using the current systems is easy to tag a single photo, and even tagging a part of a photo, like a face, has become common in sites like Flickr and Facebook. On the other hand, tagging a video sequence is more complicated and time consuming, so that users just tag the overall content of a video. In this paper we present a method for automatic video annotation that increases the number of tags originally provided by users, and localizes them temporally, associating tags to keyframes. Our approach exploits collective knowledge embedded in user-generated tags and web sources, and visual similarity of keyframes and images uploaded to social sites like YouTube and Flickr, as well as web sources like Google and Bing. Given a keyframe, our method is able to select on the fly from these visual sources the training exemplars that should be the most relevant for this test sample, and proceeds to transfer labels across similar images. Compared to existing video tagging approaches that require training classifiers for each tag, our system has few parameters, is easy to implement and can deal with an open vocabulary scenario. We demonstrate the approach on tag refinement and localization on DUT-WEBV, a large dataset of web videos, and show state-of-the-art results.