Search

Van der Waals (vdW) moires offer tunable superlattices that can strongly manipulate electronic properties. We demonstrate the in-situ manipulation of moire superlattices via heterostrain control in a vdW device. By straining a graphene layer relative to its hexagonal boron nitride substrate, we modify the shape and size of the moire. Our sliding-based technique achieves uniaxial heterostrain values exceeding 1%, resulting in distorted moires that are larger than those achievable without strain. The stretched moire is evident in transport measurements, resulting in shifted superlattice resistance peaks and Landau fans consistent with an enlarged superlattice unit cell. Electronic structure calculations reveal how heterostrain shrinks and distorts the moire Brillouin zone, resulting in a reduced electronic bandwidth as well as the appearance of highly anisotropic and quasi-1-dimensional Fermi surfaces. Our heterostrain control approach opens a wide parameter space of moire lattices to explore beyond what is possible by twist angle control alone.

Leveraging the capabilities of Knowledge Distillation (KD) strategies, we devise a strategy to fight the recent retraction of face recognition datasets. Given a pretrained Teacher model trained on a real dataset, we show that carefully utilising synthetic datasets, or a mix between real and synthetic datasets to distil knowledge from this teacher to smaller students can yield surprising results. In this sense, we trained 33 different models with and without KD, on different datasets, with different architectures and losses. And our findings are consistent, using KD leads to performance gains across all ethnicities and decreased bias. In addition, it helps to mitigate the performance gap between real and synthetic datasets. This approach addresses the limitations of synthetic data training, improving both the accuracy and fairness of face recognition models., Comment: Accepted at ECCV 2024 Workshops

As in school, one teacher to cover all subjects is insufficient to distill equally robust information to a student. Hence, each subject is taught by a highly specialised teacher. Following a similar philosophy, we propose a multiple specialized teacher framework to distill knowledge to a student network. In our approach, directed at face recognition use cases, we train four teachers on one specific ethnicity, leading to four highly specialized and biased teachers. Our strategy learns a project of these four teachers into a common space and distill that information to a student network. Our results highlighted increased performance and reduced bias for all our experiments. In addition, we further show that having biased/specialized teachers is crucial by showing that our approach achieves better results than when knowledge is distilled from four teachers trained on balanced datasets. Our approach represents a step forward to the understanding of the importance of ethnicity-specific features., Comment: Accepted at ECCV 2024 ABAW

This study investigates the effects of occlusions on the fairness of face recognition systems, particularly focusing on demographic biases. Using the Racial Faces in the Wild (RFW) dataset and synthetically added realistic occlusions, we evaluate their effect on the performance of face recognition models trained on the BUPT-Balanced and BUPT-GlobalFace datasets. We note increases in the dispersion of FMR, FNMR, and accuracy alongside decreases in fairness according to Equilized Odds, Demographic Parity, STD of Accuracy, and Fairness Discrepancy Rate. Additionally, we utilize a pixel attribution method to understand the importance of occlusions in model predictions, proposing a new metric, Face Occlusion Impact Ratio (FOIR), that quantifies the extent to which occlusions affect model performance across different demographic groups. Our results indicate that occlusions exacerbate existing demographic biases, with models placing higher importance on occlusions in an unequal fashion, particularly affecting African individuals more severely., Comment: Accepted at ECCV Workshop FAILED

Laplacian learning method is a well-established technique in classical graph-based semi-supervised learning, but its potential in the quantum domain remains largely unexplored. This study investigates the performance of the Laplacian-based Quantum Semi-Supervised Learning (QSSL) method across four benchmark datasets -- Iris, Wine, Breast Cancer Wisconsin, and Heart Disease. Further analysis explores the impact of increasing Qubit counts, revealing that adding more Qubits to a quantum system doesn't always improve performance. The effectiveness of additional Qubits depends on the quantum algorithm and how well it matches the dataset. Additionally, we examine the effects of varying entangling layers on entanglement entropy and test accuracy. The performance of Laplacian learning is highly dependent on the number of entangling layers, with optimal configurations varying across different datasets. Typically, moderate levels of entanglement offer the best balance between model complexity and generalization capabilities. These observations highlight the crucial need for precise hyperparameter tuning tailored to each dataset to achieve optimal performance in Laplacian learning methods.

This research explores the trainability of Parameterized Quantum circuit-based policies in Reinforcement Learning, an area that has recently seen a surge in empirical exploration. While some studies suggest improved sample complexity using quantum gradient estimation, the efficient trainability of these policies remains an open question. Our findings reveal significant challenges, including standard Barren Plateaus with exponentially small gradients and gradient explosion. These phenomena depend on the type of basis-state partitioning and mapping these partitions onto actions. For a polynomial number of actions, a trainable window can be ensured with a polynomial number of measurements if a contiguous-like partitioning of basis-states is employed. These results are empirically validated in a multi-armed bandit environment.

Range-based localization is ubiquitous: global navigation satellite systems (GNSS) power mobile phone-based navigation, and autonomous mobile robots can use range measurements from a variety of modalities including sonar, radar, and even WiFi signals. Many of these localization systems rely on fixed anchors or beacons with known positions acting as transmitters or receivers. In this work, we answer a fundamental question: given a set of positions we would like to localize, how should beacons be placed so as to minimize localization error? Specifically, we present an information theoretic method for optimally selecting an arrangement consisting of a few beacons from a large set of candidate positions. By formulating localization as maximum a posteriori (MAP) estimation, we can cast beacon arrangement as a submodular set function maximization problem. This approach is probabilistically rigorous, simple to implement, and extremely flexible. Furthermore, we prove that the submodular structure of our problem formulation ensures that a greedy algorithm for beacon arrangement has suboptimality guarantees. We compare our method with a number of benchmarks on simulated data and release an open source Python implementation of our algorithm and experiments., Comment: Submitted to Conference on Robots and Vision 2024

Face recognition applications have grown in parallel with the size of datasets, complexity of deep learning models and computational power. However, while deep learning models evolve to become more capable and computational power keeps increasing, the datasets available are being retracted and removed from public access. Privacy and ethical concerns are relevant topics within these domains. Through generative artificial intelligence, researchers have put efforts into the development of completely synthetic datasets that can be used to train face recognition systems. Nonetheless, the recent advances have not been sufficient to achieve performance comparable to the state-of-the-art models trained on real data. To study the drift between the performance of models trained on real and synthetic datasets, we leverage a massive attribute classifier (MAC) to create annotations for four datasets: two real and two synthetic. From these annotations, we conduct studies on the distribution of each attribute within all four datasets. Additionally, we further inspect the differences between real and synthetic datasets on the attribute set. When comparing through the Kullback-Leibler divergence we have found differences between real and synthetic samples. Interestingly enough, we have verified that while real samples suffice to explain the synthetic distribution, the opposite could not be further from being true., Comment: Accepted at FG 2024

Synthetic data is gaining increasing relevance for training machine learning models. This is mainly motivated due to several factors such as the lack of real data and intra-class variability, time and errors produced in manual labeling, and in some cases privacy concerns, among others. This paper presents an overview of the 2nd edition of the Face Recognition Challenge in the Era of Synthetic Data (FRCSyn) organized at CVPR 2024. FRCSyn aims to investigate the use of synthetic data in face recognition to address current technological limitations, including data privacy concerns, demographic biases, generalization to novel scenarios, and performance constraints in challenging situations such as aging, pose variations, and occlusions. Unlike the 1st edition, in which synthetic data from DCFace and GANDiffFace methods was only allowed to train face recognition systems, in this 2nd edition we propose new sub-tasks that allow participants to explore novel face generative methods. The outcomes of the 2nd FRCSyn Challenge, along with the proposed experimental protocol and benchmarking contribute significantly to the application of synthetic data to face recognition., Comment: arXiv admin note: text overlap with arXiv:2311.10476

Microbial contamination poses a threat to both the preservation of library and archival collections and the health of staff and users. This study investigated the microbial communities and potential health risks associated with the UNESCO-classified Norwegian Sea Trade Archive (NSTA) collection exhibiting visible microbial colonization and staff health concerns. Dust samples from book surfaces and the storage environment were analysed using culturing methods, qPCR, Next Generation Sequencing, and mycotoxin, cytotoxicity and azole resistance assays. Penicillium sp., Aspergillus sp., and Cladosporium sp. were the most common fungi identified, with some potentially toxic species like Stachybotrys sp., Toxicladosporium sp. and Aspergillus section Fumigati. Fungal resistance to azoles was not detected. Only one mycotoxin, sterigmatocystin, was found in a heavily contaminated book. Dust extracts from books exhibited moderate to high cytotoxicity on human lung cells, suggesting a potential respiratory risk. The collection had higher contamination levels compared to the storage environment, likely due to improved storage conditions. Even though, overall low contamination levels were obtained, which might be underestimated due to the presence of salt (from cod preservation) that could have interfered with the analyses. This study underlines the importance of monitoring microbial communities and implementing proper storage measures to safeguard cultural heritage and staff well-being., Comment: 17 pages, 7 figures, 2 tables

Quantum Kernels are projected to provide early-stage usefulness for quantum machine learning. However, highly sophisticated classical models are hard to surpass without losing interpretability, particularly when vast datasets can be exploited. Nonetheless, classical models struggle once data is scarce and skewed. Quantum feature spaces are projected to find better links between data features and the target class to be predicted even in such challenging scenarios and most importantly, enhanced generalization capabilities. In this work, we propose a novel approach called Systemic Quantum Score (SQS) and provide preliminary results indicating potential advantage over purely classical models in a production grade use case for the Finance sector. SQS shows in our specific study an increased capacity to extract patterns out of fewer data points as well as improved performance over data-hungry algorithms such as XGBoost, providing advantage in a competitive market as it is the FinTech and Neobank regime., Comment: Preprint

Traditionally, European social policies have focused on material well-being and social justice, neglecting subjective indicators. This review systematically examines the scientific understanding of well-being, its indicators, and its relationship with governance. It suggests that political systems and institutions significantly impact well-being, and that subjective indicators should be incorporated into public policy decisions. The findings advocate for a more holistic approach to well-being measurement, encompassing both objective and subjective dimensions., Comment: 23 pages, 4 figures, 5 tables

Reinforcement Learning (RL) consists of designing agents that make intelligent decisions without human supervision. When used alongside function approximators such as Neural Networks (NNs), RL is capable of solving extremely complex problems. Deep Q-Learning, a RL algorithm that uses Deep NNs, achieved super-human performance in some specific tasks. Nonetheless, it is also possible to use Variational Quantum Circuits (VQCs) as function approximators in RL algorithms. This work empirically studies the performance and trainability of such VQC-based Deep Q-Learning models in classic control benchmark environments. More specifically, we research how data re-uploading affects both these metrics. We show that the magnitude and the variance of the gradients of these models remain substantial throughout training due to the moving targets of Deep Q-Learning. Moreover, we empirically show that increasing the number of qubits does not lead to an exponential vanishing behavior of the magnitude and variance of the gradients for a PQC approximating a 2-design, unlike what was expected due to the Barren Plateau Phenomenon. This hints at the possibility of VQCs being specially adequate for being used as function approximators in such a context., Comment: 21 pages, 8 figures

In graphene devices, the electronic drift velocity can easily exceed the speed of sound in the material at moderate current biases. Under this condition, the electronic system can efficiently amplify acoustic phonons, leading to the exponential growth of sound waves in the direction of the carrier flow. Here, we demonstrate that such phonon amplification can significantly modify the electrical properties of graphene devices. We observe a super-linear growth of the resistivity in the direction of the carrier flow when the drift velocity exceeds the speed of sound, causing up to a 7 times increase over 8 micrometers. The resistance growth is observable for carrier densities away from the Dirac point and is enhanced at cryogenic temperatures. These observations are explained by a theoretical model for the electrical-amplification of acoustic phonons, which reach frequencies up to 2.2 terahertz with the nanoscale wavelength set by gate-tunable ~kF transitions across the Fermi surface. These findings offer a route to high-frequency on-chip sound generation and detection, which can be used to modulate and probe electronic physics in van der Waals heterostructures in the terahertz frequency range.

With the ever-growing complexity of deep learning models for face recognition, it becomes hard to deploy these systems in real life. Researchers have two options: 1) use smaller models; 2) compress their current models. Since the usage of smaller models might lead to concerning biases, compression gains relevance. However, compressing might be also responsible for an increase in the bias of the final model. We investigate the overall performance, the performance on each ethnicity subgroup and the racial bias of a State-of-the-Art quantization approach when used with synthetic and real data. This analysis provides a few more details on potential benefits of performing quantization with synthetic data, for instance, the reduction of biases on the majority of test scenarios. We tested five distinct architectures and three different training datasets. The models were evaluated on a fourth dataset which was collected to infer and compare the performance of face recognition models on different ethnicity., Comment: Accepted for Oral at BIOSIG 2023

In this article, we give a representation for compact operators acting between reflexive Banach spaces, which generalizes the representation given by Edmunds et al. for compact operators between reflexive Banach spaces with strictly convex duals. Further, we give a representation for operators on Banach spaces that are comparable to compact normal operators on Hilbert spaces and illustrate our result with an example., Comment: 16 Pages, submitted to a journal. Comments/suggestions are welcome

In recent years, advances in deep learning have resulted in a plethora of successes in the use of reinforcement learning (RL) to solve complex sequential decision tasks with high-dimensional inputs. However, existing systems lack the necessary mechanisms to provide humans with a holistic view of their competence, presenting an impediment to their adoption, particularly in critical applications where the decisions an agent makes can have significant consequences. Yet, existing RL-based systems are essentially competency-unaware in that they lack the necessary interpretation mechanisms to allow human operators to have an insightful, holistic view of their competency. Towards more explainable Deep RL (xDRL), we propose a new framework based on analyses of interestingness. Our tool provides various measures of RL agent competence stemming from interestingness analysis and is applicable to a wide range of RL algorithms, natively supporting the popular RLLib toolkit. We showcase the use of our framework by applying the proposed pipeline in a set of scenarios of varying complexity. We empirically assess the capability of the approach in identifying agent behavior patterns and competency-controlling conditions, and the task elements mostly responsible for an agent's competence, based on global and local analyses of interestingness. Overall, we show that our framework can provide agent designers with insights about RL agent competence, both their capabilities and limitations, enabling more informed decisions about interventions, additional training, and other interactions in collaborative human-machine settings., Comment: To be published in the Proceedings of the 1st World Conference on eXplainable Artificial Intelligence (xAI 2023). arXiv admin note: substantial text overlap with arXiv:2211.06376

Morphing attacks keep threatening biometric systems, especially face recognition systems. Over time they have become simpler to perform and more realistic, as such, the usage of deep learning systems to detect these attacks has grown. At the same time, there is a constant concern regarding the lack of interpretability of deep learning models. Balancing performance and interpretability has been a difficult task for scientists. However, by leveraging domain information and proving some constraints, we have been able to develop IDistill, an interpretable method with state-of-the-art performance that provides information on both the identity separation on morph samples and their contribution to the final prediction. The domain information is learnt by an autoencoder and distilled to a classifier system in order to teach it to separate identity information. When compared to other methods in the literature it outperforms them in three out of five databases and is competitive in the remaining., Comment: Accepted at EUSIPCO 2023