Thesis Tide - Find relevant and interesting papers

libcll: an Extendable Python Toolkit for Complementary-Label Learning

Complementary-label learning (CLL) is a weakly supervised learning paradigm for multiclass classification, where only complementary labels -- indicating classes an instance does not belong to -- are p...

Useful Fields:

This article presents a novel toolkit, 'libcll', aimed at tackling significant challenges in the emerging field of complementary-label learning (CLL). The systematic approach to provide a universal interface and comprehensive evaluation tools enhances its methodological rigor. It addresses practical barriers, making it highly applicable for researchers. The implications for standardization and potential for broader application within the field significantly boost its impact.

Building Trust: Foundations of Security, Safety and Transparency in AI

This paper explores the rapidly evolving ecosystem of publicly available AI models, and their potential implications on the security and safety landscape. As AI models become increasingly prevalent, u...

Useful Fields:

This article addresses critical issues related to security and safety in the context of AI models, an area that has gained increasing attention due to high-profile vulnerabilities and incidents. The proposed strategies for enhancing security and transparency are both timely and relevant, aiming for standardization in a rapidly evolving field. The rigor in reviewing existing scenarios and identifying challenges indicates a solid foundation for further research and application in AI safety protocols, which is crucial for future development in AI technologies.

A Review on Generative AI Models for Synthetic Medical Text, Time Series, and Longitudinal Data

This paper presents the results of a novel scoping review on the practical models for generating three different types of synthetic health records (SHRs): medical text, time series, and longitudinal d...

Useful Fields:

This review provides a comprehensive scoping of generative AI models applied to synthetic health records, highlighting critical areas of research and gaps. The focus on privacy preservation and the utility of various models demonstrates methodological rigor and significant applicability to the healthcare field.

DGSNA: prompt-based Dynamic Generative Scene-based Noise Addition method

This paper addresses the challenges of accurately enumerating and describing scenes and the labor-intensive process required to replicate acoustic environments using non-generative methods. We introdu...

Useful Fields:

The DGSNA method presents a novel approach to addressing existing challenges in acoustic scene simulation, utilizing prompt-based generative models in an innovative manner. Its interdisciplinary nature, combining audio engineering, AI, and scene representation, makes it highly relevant for future research and practical applications. The methodological rigor shown through comprehensive evaluations further supports its impact.

Acquire Precise and Comparable Fundus Image Quality Score: FTHNet and FQS Dataset

The retinal fundus images are utilized extensively in the diagnosis, and their quality can directly affect the diagnosis results. However, due to the insufficient dataset and algorithm application, cu...

Useful Fields:

This article demonstrates significant novelty with its establishment of a new dataset specifically aimed at improving fundus image quality assessment, which is critical for ophthalmology. The proposed FTHNet model shows methodological rigor and empirical validation, enhancing the reliability of FIQA results. The high performance metrics (PLCC and SRCC) indicate that the model is robust and applicable in real-world medical contexts, addressing current limitations in the field. The interdisciplinary nature of medical imaging and AI further underlines its relevance.

Superposition of interacting stochastic processes with memory and its application to biological data

Stochastic processes with long memories, known as long memory processes, are ubiquitous in various science and engineering problems. Superposing Markovian stochastic processes generates a non-Markovia...

Useful Fields:

This article presents a novel approach to modeling long memory stochastic processes through the superposition of interacting processes, which is a significant advancement in the field. The application to biological data, particularly migrating fish counts, showcases its relevance in real-world scenarios. The methodological rigor and the attempt to unify different models contribute to its potential impact, though further validation across different biological systems would enhance its robustness.

SMT-Layout: A MaxSMT-based Approach Supporting Real-time Interaction of Real-world GUI Layout

Leveraging the flexible expressive ability of (Max)SMT and the powerful solving ability of SMT solvers, we propose a novel layout model named SMT-Layout. SMT-Layout is the first constraint-based layou...

Useful Fields:

The article presents a novel approach to GUI layout design using MaxSMT, which is a significant advancement in the field due to its focus on real-time interaction, adaptability to various screen sizes, and the introduction of Boolean variables for hierarchy encoding. The methodology appears methodologically rigorous as it integrates advanced solving techniques which are significant for practical applications. Its implications for software design and usability make it potentially impactful, though further validation in diverse applications could enhance its robustness further.

KDC-MAE: Knowledge Distilled Contrastive Mask Auto-Encoder

In this work, we attempted to extend the thought and showcase a way forward for the Self-supervised Learning (SSL) learning paradigm by combining contrastive learning, self-distillation (knowledge dis...

Useful Fields:

The article introduces a novel architecture that effectively synthesizes multiple self-supervised learning strategies, showcasing methodological rigor through experimental results. Its interdisciplinary approach addresses key challenges in learning representations, which significantly enhances its potential impact in the field of machine learning.

Investigating the asymmetry of young stellar outflows: Combined MUSE-X-shooter study of the Th 28 jet

Characterising stellar jet asymmetries is key to setting robust constraints on jet launching models and improving our understanding of the underlying mechanisms behind jet launching. We aim to charact...

Useful Fields:

This article presents a nuanced investigation into the asymmetrical properties of stellar jets, employing sophisticated observational techniques that merge integral field spectroscopy and high-resolution spectral data. The findings regarding the differences in electron temperature, ionisation fraction, and densities provide critical insights that can challenge or refine existing models of jet launching. Its methodological rigor and clear implications for mass accretion dynamics enhance its relevance.

A Neural Denoising Vocoder for Clean Waveform Generation from Noisy Mel-Spectrogram based on Amplitude and Phase Predictions

This paper proposes a novel neural denoising vocoder that can generate clean speech waveforms from noisy mel-spectrograms. The proposed neural denoising vocoder consists of two components, i.e., a spe...

Useful Fields:

The article presents a novel approach to speech synthesis that significantly enhances the quality of generated waveforms from noisy inputs, using advanced neural network architectures. The state-of-the-art performance reported on established datasets indicates not only methodological rigor but also the potential for practical applications in various domains. Its innovative combination of spectrum and enhancement modules captures both the amplitude and phase information that are typically challenging in vocoder designs, which establishes a strong foundation for future developments in speech technology.

Cost of controllability of the Burgers equation linearized at a steady shock in the vanishing viscosity limit

We consider the one-dimensional Burgers equation linearized at a stationary shock, and investigate its null-controllability cost with a control at the left endpoint. We give an upper and a lower bound...

Useful Fields:

The article presents a novel investigation into the controllability of the Burgers equation, particularly at the nuanced transition of the vanishing viscosity limit. This focus on null-controllability in a specialized context is likely to yield significant insights into the behavior of controlled systems, making it important for advancing theoretical understanding and application in fluid dynamics and control theory. The use of complex analysis and the adaptation of existing methods suggests methodological rigor, while the exploration of boundaries in control time further strengthens its relevance.

Comparing design and off-design aerodynamic performance of a natural laminar airfoil

Natural laminar flow airfoils are essential technologies designed to reduce drag and significantly enhance aerodynamic performance. A notable example is the SHM1 airfoil, created to meet the requireme...

Useful Fields:

The article presents a comprehensive comparison of the aerodynamic performance of a natural laminar flow airfoil, discussing both design and off-design operational conditions. The methodology includes extensive empirical testing and advanced numerical simulations, contributing to a nuanced understanding of flow dynamics. The study's findings on shock interactions and drag-divergence behavior are highly relevant for aerospace engineering, particularly for the development of more efficient aircraft designs. Its application to real-world scenarios, such as business jets, enhances its significance in the field.

On the Accuracy and Precision of Moving Averages to Estimate Wi-Fi Link Quality

The radio spectrum is characterized by a noticeable variability, which impairs performance and determinism of every wireless communication technology. To counteract this aspect, mechanisms like Minstr...

Useful Fields:

The work addresses a pertinent issue in wireless communication by critically analyzing the effectiveness of moving averages in estimating Wi-Fi link quality. Its relevance increases due to its implications for machine learning applications in networking that aim to optimize link quality. However, while the techniques explored are foundational, they may not offer groundbreaking insights into novel methodologies or innovative solutions.

The Product of linear forms over function fields

The aim of this paper is to study the product of $n$ linear forms over function fields. We calculate the maximum value of the minima of the forms with determinant one when $n$ is small...

Useful Fields:

This article presents a significant theoretical advancement in the study of function fields by deriving new results about the maxima of minima of linear forms. It employs innovative methodologies, such as reduction theory, and connects algebraic results to periodic dynamics, which demonstrates interdisciplinary depth. The findings could robustly impact future research in algebra, number theory, and dynamical systems.

PowerMove: Optimizing Compilation for Neutral Atom Quantum Computers with Zoned Architecture

Neutral atom-based quantum computers (NAQCs) have recently emerged as promising candidates for scalable quantum computing, largely due to their advanced hardware capabilities, particularly qubit movem...

Useful Fields:

The paper presents a highly novel and rigorous approach to compiler optimization for a cutting-edge quantum computing architecture. The advancements in fidelity and execution time are substantial, indicating practical utility and the potential for significant impact in the field. Moreover, the promise of open-sourcing the implementation encourages community collaboration, enhancing its relevance.

Low-resource Machine Translation: what for? who for? An observational study on a dedicated Tetun language translation service

The impact of machine translation (MT) on low-resource languages remains poorly understood. In particular, observational studies of actual usage patterns are scarce. Such studies could provide valuabl...

Useful Fields:

This article presents a novel observational study that fills a critical gap in understanding how machine translation is used in low-resource languages. Its methodological rigor is notable, utilizing actual usage data instead of relying solely on surveys. The practical implications for the design of MT systems that serve marginalized language communities make it particularly impactful. The study’s conclusions can profoundly influence the development of localized MT services, signifying high relevance for both academic and practical applications.

Stripe Antiferromagnetic Ground State of FeSe Revealed by Density Functional Theory

The magnetic ground state of iron selenide (FeSe) has been a topic of debate, with experimental evidence suggesting stripe spin fluctuations as predominant at low temperatures, while density functiona...

Useful Fields:

The article provides a novel approach to investigating the magnetic properties of FeSe using an advanced DFT functional, demonstrating that the $ ext{r}^{2} ext{SCAN}$ functional may reveal new magnetic states that traditional methods failed to capture. This contributes to an ongoing debate in condensed matter physics and may influence future research directions in similar materials, making it highly relevant in the field.

Noncollinear ferroelectric and screw-type antiferroelectric phases in a metal-free hybrid molecular crystal

Noncollinear dipole textures greatly extend the scientific merits and application perspective of ferroic materials. In fact, noncollinear spin textures have been well recognized as one of the core iss...

Useful Fields:

This article presents significant findings related to noncollinear dipole textures in ferroelectric and antiferroelectric materials, an area previously underexplored. The combination of experimental characterizations and ab initio calculations enhances the methodological rigor and provides a strong basis for the claims made. Furthermore, the identification of a unique transition mechanism could open new avenues for research in ferroelectric materials and their applications, particularly in devices exploiting noncollinear polarities.

Prototype Optimization with Neural ODE for Few-Shot Learning

Few-Shot Learning (FSL) is a challenging task, which aims to recognize novel classes with few examples. Pre-training based methods effectively tackle the problem by pre-training a feature extractor an...

Useful Fields:

This article presents a novel approach to prototype optimization in Few-Shot Learning, addressing critical issues such as prototype bias and gradient bias in sparse data scenarios. The introduction of a Neural ODE-based meta-optimizer is particularly innovative, and the proposed solutions demonstrate strong empirical results. The combination of novelty and methodological rigor supports its potential impact in the field.

E-STGCN: Extreme Spatiotemporal Graph Convolutional Networks for Air Quality Forecasting

Modeling and forecasting air quality plays a crucial role in informed air pollution management and protecting public health. The air quality data of a region, collected through various pollution monit...

Useful Fields:

The article presents a novel approach in air quality forecasting by integrating extreme value theory with spatiotemporal graph convolutional networks (E-STGCN). This combination addresses key challenges in predicting extreme air pollution events, which is significant for public health. The methodological rigor is evident from its application to real-world data across multiple locations, showcasing both robustness and practical applicability. The focus on extreme air pollutant levels fills a critical gap in existing research, making it highly impactful for environmental monitoring and management.