Search

This paper evaluates the use of metamorphic relations to enhance the robustness and real-world performance of machine learning models. We propose a Metamorphic Retraining Framework, which applies metamorphic relations to data and utilizes semi-supervised learning algorithms in an iterative and adaptive multi-cycle process. The framework integrates multiple semi-supervised retraining algorithms, including FixMatch, FlexMatch, MixMatch, and FullMatch, to automate the retraining, evaluation, and testing of models with specified configurations. To assess the effectiveness of this approach, we conducted experiments on CIFAR-10, CIFAR-100, and MNIST datasets using a variety of image processing models, both pretrained and non-pretrained. Our results demonstrate the potential of metamorphic retraining to significantly improve model robustness as we show in our results that each model witnessed an increase of an additional flat 17 percent on average in our robustness metric.

The functionality of magnetoelectric multiferroics depends on the formation, size, and coupling of their magnetic and electric domains. Knowing the parameters guiding these criteria is a key effort in the emerging field of magnetoelectric domain engineering. Here we show, using a combination of piezoresponse-force microscopy, non-linear optics, and x-ray scattering, that the correlation length setting the size of the ferroelectric domains in the multiferroic hexagonal manganites can be engineered from the micron range down to a few unit cells under the substitution of Mn$^{3+}$ ions with Al$^{3+}$ ions. The magnetoelectric coupling mechanism between the antiferromagnetic Mn$^{3+}$ order and the distortive-ferroelectric order remains intact even at substantial replacement of Mn$^{3+}$ by Al$^{3+}$. Hence, chemical substitution proves to be an effective tool for domain-size engineering in one of the most studied classes of multiferroics., Comment: 10 pages, 8 figures

Segmenting the boundary between tumor and healthy tissue during surgical cancer resection poses a significant challenge. In recent years, Hyperspectral Imaging (HSI) combined with Machine Learning (ML) has emerged as a promising solution. However, due to the extensive information contained within the spectral domain, most ML approaches primarily classify individual HSI (super-)pixels, or tiles, without taking into account their spatial context. In this paper, we propose an improved methodology that leverages the spatial context of tiles for more robust and smoother segmentation. To address the irregular shapes of tiles, we utilize Graph Neural Networks (GNNs) to propagate context information across neighboring regions. The features for each tile within the graph are extracted using a Convolutional Neural Network (CNN), which is trained simultaneously with the subsequent GNN. Moreover, we incorporate local image quality metrics into the loss function to enhance the training procedure's robustness against low-quality regions in the training images. We demonstrate the superiority of our proposed method using a clinical ex vivo dataset consisting of 51 HSI images from 30 patients. Despite the limited dataset, the GNN-based model significantly outperforms context-agnostic approaches, accurately distinguishing between healthy and tumor tissues, even in images from previously unseen patients. Furthermore, we show that our carefully designed loss function, accounting for local image quality, results in additional improvements. Our findings demonstrate that context-aware GNN algorithms can robustly find tumor demarcations on HSI images, ultimately contributing to better surgery success and patient outcome., Comment: 11 pages, 6 figures

Ferroelectric domain wall (DW) conductivity (DWC) can be attributed to two separate mechanisms: (a) the injection/ejection of charge carriers across the Schottky barrier formed at the (metal-) electrode-DW junction and (b) the transport of those charge carriers along the DW. Current-voltage (IU) characteristics, recorded at variable temperatures from LiNbO$_3$ (LNO) DWs, are clearly able to differentiate between these two contributions. Practically, they allow us here to directly quantify the physical parameters relevant for the two mechanisms (a) and (b) mentioned above. These are, e.g., the resistance of the DW, the saturation current, the ideality factor, and the Schottky barrier height of the electrode/DW junction. Furthermore, the activation energies needed to initiate the thermally-activated electronic transport along the DWs, can be extracted. In addition, we show that electronic transport along LiNbO$_3$ DWs can be elegantly viewed and interpreted in an adapted semiconductor picture based on a double-diode/double-resistor equivalent circuit model, the R2D2 model. Finally, our R2D2 model was checked for its universality by fitting the DWC data not only to z-cut LNO bulk DWs, but equally to z-cut thin-film LNO DWs, and DWC from x-cut DWs as reported in literature., Comment: 10 pages, 7 figures; with included supporting information (6 pages, 4 figures)

When we speak, the prosody and content of the speech can be inferred from the movement of our lips. In this work, we explore the task of lip to speech synthesis, i.e., learning to generate speech given only the lip movements of a speaker where we focus on learning accurate lip to speech mappings for multiple speakers in unconstrained, large vocabulary settings. We capture the speaker's voice identity through their facial characteristics, i.e., age, gender, ethnicity and condition them along with the lip movements to generate speaker identity aware speech. To this end, we present a novel method "Lip2Speech", with key design choices to achieve accurate lip to speech synthesis in unconstrained scenarios. We also perform various experiments and extensive evaluation using quantitative, qualitative metrics and human evaluation.

Attacks on machine learning models have been, since their conception, a very persistent and evasive issue resembling an endless cat-and-mouse game. One major variant of such attacks is poisoning attacks which can indirectly manipulate an ML model. It has been observed over the years that the majority of proposed effective defense models are only effective when an attacker is not aware of them being employed. In this paper, we show that the attack-agnostic De-Pois defense is hardly an exception to that rule. In fact, we demonstrate its vulnerability to the simplest White-Box and Black-Box attacks by an attacker that knows the structure of the De-Pois defense model. In essence, the De-Pois defense relies on a critic model that can be used to detect poisoned data before passing it to the target model. In our work, we break this poison-protection layer by replicating the critic model and then performing a composed gradient-sign attack on both the critic and target models simultaneously -- allowing us to bypass the critic firewall to poison the target model.

Although GPS has been considered a ubiquitous outdoor localization technology, we are still far from a similar technology for indoor environments. While a number of technologies have been proposed for indoor localization, they are isolated efforts that are way from a true ubiquitous localization system. A ubiquitous indoor positioning system is envisioned to be deployed on a large scale worldwide, with minimum overhead, to work with heterogeneous devices, and to allow users to roam seamlessly from indoor to outdoor environments. Such a system will enable a wide set of applications including worldwide seamless direction finding between indoor locations, enhancing first responders' safety by providing anywhere localization and floor plans, and providing a richer environment for location-aware social networking applications. We describe an architecture for the ubiquitous indoor positioning system (IPS) and the challenges that have to be addressed to materialize it. We then focus on the feasibility of automating the construction of a worldwide indoor floor plan and fingerprint database which, as we believe, is one of the main challenges that limit the existence of a ubiquitous IPS system. Our proof of concept uses a crowd-sourcing approach that leverages the embedded sensors in today's cell phones as a worldwide distributed floor plan generation tool. This includes constructing the floor plans and determining the areas of interest (corridors, offices, meeting rooms, elevators, etc). The cloud computing concepts are also adopted for the processing and storage of the huge amount of data generated and requested by the system's users. Our results show the ability of the system to construct an accurate floor plan and identify the areas of interest with more than 90% accuracy. We also identify different research directions for addressing the challenges of realizing a true ubiquitous IPS system., Comment: Under submission

The decay of inner-shell vacancy in an atom through radiative and non-radiative transitions leads to final charged ions. The de-excitation decay of 3s, 3p and 3d vacancies in Kr atoms are calculated using Monte-Carlo simulation method. The vacancy cascade pathway resulted from the de-excitation decay of deep core hole in 3s subshell in Kr atoms is discussed. The generation of spectator vacancies during the vacancy cascade development gives rise to Auger satellite spectra. The last transitions of the de-excitation decay of 3s, 3p and 3d holes lead to specific charged ions. Dirac-Fock-Slater wave functions are adapted to calculate radiative and non-radiative transition probabilities. The intensity of Kr^{4+} ions are high for 3s hole state, whereas Kr^{3+} and Kr^{2+} ions have highest intensities for 3p and 3d hole states, respectively. The present results of ion charge state distributions agree well with the experimental data., Comment: Published in SIGMA (Symmetry, Integrability and Geometry: Methods and Applications) at http://www.emis.de/journals/SIGMA/