arXiv subjects

Bin Hu

Publications and source records attributed to Bin Hu.

At least 127 records · Page 7Linked to original sources

arXiv2025

RBA-FE: A Robust Brain-Inspired Audio Feature Extractor for Depression Diagnosis

This article proposes a robust brain-inspired audio feature extractor (RBA-FE) model for depression diagnosis, using an improved hierarchical network architecture. Most deep learning models achieve state-of-the-art performance for image-based diagnostic tasks, ignoring the counterpart audio features. In order to tailor the noise challenge, RBA-FE leverages six acoustic features extracted from the raw audio, capturing both spatial characteristics and temporal dependencies. This hybrid attribute helps alleviate the precision limitation in audio feature extraction within other learning models like deep residual shrinkage networks. To deal with the noise issues, our model incorporates an improved spiking neuron model, called adaptive rate smooth leaky integrate-and-fire (ARSLIF). The ARSLIF model emulates the mechanism of ``retuning of cellular signal selectivity" in the brain attention systems, which enhances the model robustness against environmental noises in audio data. Experimental results demonstrate that RBA-FE achieves state-of-the-art accuracy on the MODMA dataset, respectively with 0.8750, 0.8974, 0.8750 and 0.8750 in precision, accuracy, recall and F1 score. Extensive experiments on the AVEC2014 and DAIC-WOZ datasets both show enhancements in noise robustness. It is further indicated by comparison that the ARSLIF neuron model suggest the abnormal firing pattern within the feature extraction on depressive audio data, offering brain-inspired interpretability.

cs.SD↗

arXiv2025

Odd-Parity Quasiparticle Interference in the Superconductive Surface State of UTe2

Although no known material exhibits intrinsic topological superconductivity, wherein spin-triplet odd-parity electron pairing occurs, UTe2 is now the leading representative of this class. Conventionally, the parity of the superconducting order parameter may be established by using Bogoliubov quasiparticle interference (QPI) imaging. However, odd-parity superconductors should support a topological quasiparticle surface band (QSB) at energies within the maximum superconducting energy gap. QPI would then be dominated by the electronic structure of the QSB and only reveal the characteristics of the bulk order parameter excursively. Here, we visualize quasiparticle interference patterns of UTe2 and find that, at the (0-11) cleave surface, a new band of Bogoliubov quasiparticles appears only in the superconducting state. QPI visualization then allows study of dispersion of states within this QSB, which we demonstrate exists only within the range of Fermi momenta projected onto the (0-11) surface. Finally, we develop a theoretical framework to predict the QPI signatures of such a QSB at the (0-11) surface of UTe2. Its predictions are most consistent with the experimental results if the bulk superconducting gap function exhibits time-reversal conserving, odd-parity, a-axis nodal, B3u symmetry.

cond-mat.supr-con↗

arXiv2025

MADUV: The 1st INTERSPEECH Mice Autism Detection via Ultrasound Vocalization Challenge

The Mice Autism Detection via Ultrasound Vocalization (MADUV) Challenge introduces the first INTERSPEECH challenge focused on detecting autism spectrum disorder (ASD) in mice through their vocalizations. Participants are tasked with developing models to automatically classify mice as either wild-type or ASD models based on recordings with a high sampling rate. Our baseline system employs a simple CNN-based classification using three different spectrogram features. Results demonstrate the feasibility of automated ASD detection, with the considered audible-range features achieving the best performance (UAR of 0.600 for segment-level and 0.625 for subject-level classification). This challenge bridges speech technology and biomedical research, offering opportunities to advance our understanding of ASD models through machine learning approaches. The findings suggest promising directions for vocalization analysis and highlight the potential value of audible and ultrasound vocalizations in ASD detection.

cs.SD↗

arXiv2025

Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching

We introduce Adjoint Sampling, a highly scalable and efficient algorithm for learning diffusion processes that sample from unnormalized densities, or energy functions. It is the first on-policy approach that allows significantly more gradient updates than the number of energy evaluations and model samples, allowing us to scale to much larger problem settings than previously explored by similar methods. Our framework is theoretically grounded in stochastic optimal control and shares the same theoretical guarantees as Adjoint Matching, being able to train without the need for corrective measures that push samples towards the target distribution. We show how to incorporate key symmetries, as well as periodic boundary conditions, for modeling molecules in both cartesian and torsional coordinates. We demonstrate the effectiveness of our approach through extensive experiments on classical energy functions, and further scale up to neural network-based energy models where we perform amortized conformer generation across many molecular systems. To encourage further research in developing highly scalable sampling methods, we plan to open source these challenging benchmarks, where successful methods can directly impact progress in computational chemistry.

cs.LG↗

arXiv2025

Hybrid Audio Detection Using Fine-Tuned Audio Spectrogram Transformers: A Dataset-Driven Evaluation of Mixed AI-Human Speech

The rapid advancement of artificial intelligence (AI) has enabled sophisticated audio generation and voice cloning technologies, posing significant security risks for applications reliant on voice authentication. While existing datasets and models primarily focus on distinguishing between human and fully synthetic speech, real-world attacks often involve audio that combines both genuine and cloned segments. To address this gap, we construct a novel hybrid audio dataset incorporating human, AI-generated, cloned, and mixed audio samples. We further propose fine-tuned Audio Spectrogram Transformer (AST)-based models tailored for detecting these complex acoustic patterns. Extensive experiments demonstrate that our approach significantly outperforms existing baselines in mixed-audio detection, achieving 97\% classification accuracy. Our findings highlight the importance of hybrid datasets and tailored models in advancing the robustness of speech-based authentication systems.

cs.SD↗

arXiv2025

Hints of Nonminimally Coupled Gravity in DESI 2024 Baryon Acoustic Oscillation Measurements

The cosmic microwave background (CMB) and baryon acoustic oscillations (BAO) are two of the most robust observations in cosmology. The recent BAO measurements from the DESI collaboration have presented, for the first time, inconsistency between BAO and CMB within the standard cosmological model $Λ$CDM, indicating a preference for dynamical dark energy over a cosmological constant. We analyze the theoretical implication of the DESI BAO observation for dark energy and gravity employing a nonparametric reconstruction approach for both the dark energy equation of state $w_{\rm DE}(a)$ and the effective field theory coefficients. We find that the DESI data can rule out quintessence dark energy by indicating a crossing of the phantom divide at $z\lesssim1$. Furthermore, when analyzed within the broad context of Horndeski gravity which includes general relativity and many known modified gravity theories such as generalized Galileons, $f(R)$ and Brans-Dicke, our result implies that gravity should be nonminimally coupled to explain the observations, establishing the DESI result as the first hint of modified gravity. Based on these insights, we propose the \textit{thawing gravity} model to explain the nonminimal coupling and phantom crossing indicated by observation, which also fits better to DESI BAO, CMB and type Ia Supernovae data than $Λ$CDM.

astro-ph.CO↗

arXiv2025

An interference-based method for the detection of strongly lensed gravitational waves

The strongly lensed gravitational wave (SLGW) is a promising transient phenomenon. However, the long-wave nature of gravitational waves poses a significant challenge in identification of its host galaxy. To tackle this challenge, we propose a method triggered by the wave optics effect of microlensing. The microlensing interference introduce frequency-dependent fluctuations in the waveform. Our method consists of three steps. First, we reconstruct the waveforms by using the template-independent and template-dependent methods. The mismatch of two reconstructions serves as an indicator of SLGWs. This step can identify approximately $10\%$ SLGWs. Second, we pair the SLGWs' multiple-images by employing the sky localization overlapping. Because we have pre-identified at least one image through microlensing, the false alarm probability for pairing SLGWs is significantly reduced. Third, we search the host galaxy by requiring the consistency of time-delays between galaxy-galaxy lensing and SLGW. By combing the stage-IV galaxy survey and the third-generation gravitational wave detectors, we expect to find, on average, 1 quadruple-image system per 3 years. The merit of this method can significantly facilitate the pursuit of time-delay cosmography, discovery of compact objects and multi-messenger astronomy.

astro-ph.IM↗

arXiv2025

DECIDER: A Dual-System Rule-Controllable Decoding Framework for Language Generation

Constrained decoding approaches aim to control the meaning or style of text generated by the pre-trained large language models (LLMs or also PLMs) for various tasks at inference time. However, these methods often guide plausible continuations by greedily and explicitly selecting targets. Though fulfilling the task requirements, these methods may overlook certain general and natural logics that humans would implicitly follow towards such targets. Inspired by cognitive dual-process theory, in this work, we propose a novel decoding framework DECIDER where the base LLMs are equipped with a First-Order Logic (FOL) reasoner to express and evaluate the rules, along with a decision function that merges the outputs of both systems to guide the generation. Unlike previous constrained decodings, DECIDER transforms the encouragement of target-specific words into all words that satisfy several high-level rules, enabling us to programmatically integrate our logic into LLMs. Experiments on CommonGen and PersonaChat demonstrate that DECIDER effectively follows given FOL rules to guide LLMs in a more human-like and logic-controlled manner.

cs.CL↗

arXiv2025

Neural Contraction Metrics with Formal Guarantees for Discrete-Time Nonlinear Dynamical Systems

Contraction metrics are crucial in control theory because they provide a powerful framework for analyzing stability, robustness, and convergence of various dynamical systems. However, identifying these metrics for complex nonlinear systems remains an open challenge due to the lack of scalable and effective tools. This paper explores the approach of learning verifiable contraction metrics parametrized as neural networks (NNs) for discrete-time nonlinear dynamical systems. While prior works on formal verification of contraction metrics for general nonlinear systems have focused on convex optimization methods (e.g. linear matrix inequalities, etc) under the assumption of continuously differentiable dynamics, the growing prevalence of NN-based controllers, often utilizing ReLU activations, introduces challenges due to the non-smooth nature of the resulting closed-loop dynamics. To bridge this gap, we establish a new sufficient condition for establishing formal neural contraction metrics for general discrete-time nonlinear systems assuming only the continuity of the dynamics. We show that from a computational perspective, our sufficient condition can be efficiently verified using the state-of-the-art neural network verifier $α,\!β$-CROWN, which scales up non-convex neural network verification via novel integration of symbolic linear bound propagation and branch-and-bound. Built upon our analysis tool, we further develop a learning method for synthesizing neural contraction metrics from sampled data. Finally, our approach is validated through the successful synthesis and verification of NN contraction metrics for various nonlinear examples.

math.OC↗

arXiv2025

Multi-messenger Gravitational Lensing

We introduce the rapidly emerging field of multi-messenger gravitational lensing - the discovery and science of gravitationally lensed phenomena in the distant universe through the combination of multiple messengers. This is framed by gravitational lensing phenomenology that has grown since the first discoveries in the 20th century, messengers that span 30 orders of magnitude in energy from high energy neutrinos to gravitational waves, and powerful "survey facilities" that are capable of continually scanning the sky for transient and variable sources. Within this context, the main focus is on discoveries and science that are feasible in the next 5-10 years with current and imminent technology including the LIGO-Virgo-KAGRA network of gravitational wave detectors, the Vera C. Rubin Observatory, and contemporaneous gamma/X-ray satellites and radio surveys. The scientific impact of even one multi-messenger gravitational lensing discovery will be transformational and reach across fundamental physics, cosmology and astrophysics. We describe these scientific opportunities and the key challenges along the path to achieving them. This article is the introduction to the Theme Issue of the Philosophical Transactions of The Royal Society A on the topic of Multi-messenger Gravitational Lensing, and describes the consensus that emerged at the associated Theo Murphy Discussion Meeting in March 2024.

astro-ph.HE↗

arXiv2025

Improved Scalable Lipschitz Bounds for Deep Neural Networks

Computing tight Lipschitz bounds for deep neural networks is crucial for analyzing their robustness and stability, but existing approaches either produce relatively conservative estimates or rely on semidefinite programming (SDP) formulations (namely the LipSDP condition) that face scalability issues. Building upon ECLipsE-Fast, the state-of-the-art Lipschitz bound method that avoids SDP formulations, we derive a new family of improved scalable Lipschitz bounds that can be combined to outperform ECLipsE-Fast. Specifically, we leverage more general parameterizations of feasible points of LipSDP to derive various closed-form Lipschitz bounds, avoiding the use of SDP solvers. In addition, we show that our technique encompasses ECLipsE-Fast as a special case and leads to a much larger class of scalable Lipschitz bounds for deep neural networks. Our empirical study shows that our bounds improve ECLipsE-Fast, further advancing the scalability and precision of Lipschitz estimation for large neural networks.

cs.LG↗

arXiv2025

DynaMath: A Dynamic Visual Benchmark for Evaluating Mathematical Reasoning Robustness of Vision Language Models

The rapid advancements in Vision-Language Models (VLMs) have shown great potential in tackling mathematical reasoning tasks that involve visual context. Unlike humans who can reliably apply solution steps to similar problems with minor modifications, we found that SOTA VLMs like GPT-4o can consistently fail in these scenarios, revealing limitations in their mathematical reasoning capabilities. In this paper, we investigate the mathematical reasoning robustness in VLMs and evaluate how well these models perform under different variants of the same question, such as changes in visual numerical values or function graphs. While several vision-based math benchmarks have been developed to assess VLMs' problem-solving capabilities, these benchmarks contain only static sets of problems and cannot easily evaluate mathematical reasoning robustness. To fill this gap, we introduce DynaMath, a dynamic visual math benchmark designed for in-depth assessment of VLMs. DynaMath includes 501 high-quality, multi-topic seed questions, each represented as a Python program. Those programs are carefully designed and annotated to enable the automatic generation of a much larger set of concrete questions, including many different types of visual and textual variations. DynaMath allows us to evaluate the generalization ability of VLMs, by assessing their performance under varying input conditions of a seed question. We evaluated 14 SOTA VLMs with 5,010 generated concrete questions. Our results show that the worst-case model accuracy, defined as the percentage of correctly answered seed questions in all 10 variants, is significantly lower than the average-case accuracy. Our analysis emphasizes the need to study the robustness of VLMs' reasoning abilities, and DynaMath provides valuable insights to guide the development of more reliable models for mathematical reasoning.

cs.CV↗

arXiv2025

Surface Vision Mamba: Leveraging Bidirectional State Space Model for Efficient Spherical Manifold Representation

Attention-based methods have demonstrated exceptional performance in modelling long-range dependencies on spherical cortical surfaces, surpassing traditional Geometric Deep Learning (GDL) models. However, their extensive inference time and high memory demands pose challenges for application to large datasets with limited computing resources. Inspired by the state space model in computer vision, we introduce the attention-free Vision Mamba (Vim) to spherical surfaces, presenting a domain-agnostic architecture for analyzing data on spherical manifolds. Our method achieves surface patching by representing spherical data as a sequence of triangular patches derived from a subdivided icosphere. The proposed Surface Vision Mamba (SiM) is evaluated on multiple neurodevelopmental phenotype regression tasks using cortical surface metrics from neonatal brains. Experimental results demonstrate that SiM outperforms both attention- and GDL-based methods, delivering 4.8 times faster inference and achieving 91.7% lower memory consumption compared to the Surface Vision Transformer (SiT) under the Ico-4 grid partitioning. Sensitivity analysis further underscores the potential of SiM to identify subtle cognitive developmental patterns. The code is available at https://github.com/Rongzhao-He/surface-vision-mamba.

cs.CV↗

arXiv2025

Progress of the TianQin project

TianQin is a future space-based gravitational wave observatory targeting the frequency window of $10^{-4}$ Hz $\sim 1$ Hz. A large variety of gravitational wave sources are expected in this frequency band, including the merger of massive black hole binaries, the inspiral of extreme/intermediate mass ratio systems, stellar-mass black hole binaries, Galactic compact binaries, and so on. TianQin will consist of three Earth orbiting satellites on nearly identical orbits with orbital radii of about $10^5$ km. The satellites will form a normal triangle constellation whose plane is nearly perpendicular to the ecliptic plane. The TianQin project has been progressing smoothly following the ``0123" technology roadmap. In step ``0", the TianQin laser ranging station has been constructed and it has successfully ranged to all the five retro-reflectors on the Moon. In step ``1", the drag-free control technology has been tested and demonstrated using the TianQin-1 satellite. In step ``2", the inter-satellite laser interferometry technology will be tested using the pair of TianQin-2 satellites. The TianQin-2 mission has been officially approved and the satellites will be launched around 2026. In step ``3", i.e., the TianQin-3 mission, three identical satellites will be launched around 2035 to form the space-based gravitational wave detector, TianQin, and to start gravitational wave detection in space.

gr-qc↗

arXiv2025

Revealing the orbital origins of exotic electronic states with Ti substitution in kagome superconductor CsV3Sb5

The multiband kagome superconductor CsV3Sb5 exhibits complex orbital textures on the Fermi surface, making the orbital origins of its cascade of correlated electronic states and superconductivity a major scientific puzzle. Chemical doping of the kagome plane can simultaneously tune the exotic states and the Fermi-surface orbital texture, and thus offers a unique opportunity to correlate the given states with specific orbitals. In this Letter, by substituting V atoms with Ti in kagome superconductor CsV3Sb5, we reveal the orbital origin of a cascade of its correlated electronic states through the orbital-resolved quasiparticle interference (QPI). We analyze the QPI changes associated with different orbitals, aided by first-principles calculations. We have observed that the in-plane and out-of-plane vanadium 3d orbitals cooperate to form unidirectional coherent states in pristine CsV3Sb5, whereas the out-of-plane component disappears with doping-induced suppression of charge density wave and global electronic nematicity. In addition, the Sb pz orbital plays an important role in both the pseudo-gap and superconducting states in CsV3Sb5. Our findings offer new insights into multiorbital physics in quantum materials which are generally manifested with intriguing correlations between atomic orbitals and symmetry-encoded correlated electronic states.

cond-mat.supr-con↗

arXiv2025

ISAM-MTL: Cross-subject multi-task learning model with identifiable spikes and associative memory networks

Cross-subject variability in EEG degrades performance of current deep learning models, limiting the development of brain-computer interface (BCI). This paper proposes ISAM-MTL, which is a multi-task learning (MTL) EEG classification model based on identifiable spiking (IS) representations and associative memory (AM) networks. The proposed model treats EEG classification of each subject as an independent task and leverages cross-subject data training to facilitate feature sharing across subjects. ISAM-MTL consists of a spiking feature extractor that captures shared features across subjects and a subject-specific bidirectional associative memory network that is trained by Hebbian learning for efficient and fast within-subject EEG classification. ISAM-MTL integrates learned spiking neural representations with bidirectional associative memory for cross-subject EEG classification. The model employs label-guided variational inference to construct identifiable spike representations, enhancing classification accuracy. Experimental results on two BCI Competition datasets demonstrate that ISAM-MTL improves the average accuracy of cross-subject EEG classification while reducing performance variability among subjects. The model further exhibits the characteristics of few-shot learning and identifiable neural activity beneath EEG, enabling rapid and interpretable calibration for BCI systems.

cs.NE↗

arXiv2025

Forecasts of effects of beam systematics and deprojection on the third-generation ground-based cosmic microwave background experiment

The ground-based cosmic microwave background (CMB) experiments are susceptible to various instrumental errors, especially for $B$-mode measurements. The difference between the response of two polarized detectors, referred to as the beam mismatch, would induce a $T\rightarrow P$ leakage when the detector pair is differenced to cancel the unpolarized signal. We applied the deprojection technique on the time-ordered mock data to mitigate the systematic contamination caused by beam mismatches by assuming the third-generation ground-based CMB experiment (S3). Our results show that the deprojection effectively recovered the input power spectra. We adopted the Needlet ILC (NILC) and constrained ILC (cILC) methods to reconstruct the foreground-cleaned $TEB$ maps, and we evaluated the level of residual systematic errors after the foreground cleaning pipeline by comparing the power spectra between the systematics-added data after deprojection and the systematics-free data. The results show that the residual beam systematics cleaned by deprojection do not bias the CMB measurements of the $T$, $E$, and $B$ modes nor the CMB lensing reconstruction or the estimation of the tensor-to-scalar ratio under the S3 sensitivity.

astro-ph.CO↗

arXiv2025

Forecast of gravitationally lensed Type Ia supernovae time delay measurement by Muztage-Ata 1.93m Synergy Telescope

Strong lensing time delay measurement is a promising method to address the Hubble tension, offering a completely independent approach compared to both the cosmic microwave background analysis and the local distance ladder. As a third-party examination of the Hubble tension, this method provides a unique perspective. Strongly lensed quasar (glQSO) systems have demonstrated significant potential in tackling this issue, achieving an impressive $2\%$ accuracy level. However, advancing to $1\%$ or sub-percent accuracy is challenging due to several intrinsic limitations of glQSOs. Fortunately, strongly lensed supernovae (glSNe) offer a more robust solution, thanks to their characteristic light curve, significant brightness variations, and additional advantages. The Muztagh-Ata 1.93m Synergy Telescope (MOST) is an exceptional instrument for monitoring strong lensing time delays. In this study, we simulate the follow-up multi-band light curve monitoring for glSNe Ia systems, which are expected to be firstly discovered by the Chinese Survey Space Telescope (CSST). Our results show that with $300s \times 9$ exposures in each epoch, MOST can achieve a signal-to-noise ratio (SNR) of approximately 50 for the brightest images of glSNe Ia, while even the faintest images maintain an SNR of at least 7. Using a standard SNe Ia light curve template for fitting, we measured the time delays. With a 2-day cadence, MOST achieves a time delay error of only a few hours, with the bias typically remaining below one hour. This study highlights the capability of MOST to significantly advance the precision of time delay measurements, offering a promising path toward resolving the Hubble tension.

astro-ph.GA↗