arXiv subjects

Meng Wang

Publications and source records attributed to Meng Wang.

At least 325 records · Page 18Linked to original sources

arXiv2025

Knowledge Swapping via Learning and Unlearning

We introduce \textbf{Knowledge Swapping}, a novel task designed to selectively regulate knowledge of a pretrained model by enabling the forgetting of user\-specified information, retaining essential knowledge, and acquiring new knowledge simultaneously. By delving into the analysis of knock-on feature hierarchy, we find that incremental learning typically progresses from low\-level representations to higher\-level semantics, whereas forgetting tends to occur in the opposite direction\-starting from high-level semantics and moving down to low-level features. Building upon this, we propose to benchmark the knowledge swapping task with the strategy of \textit{Learning Before Forgetting}. Comprehensive experiments on various tasks like image classification, object detection, and semantic segmentation validate the effectiveness of the proposed strategy. The source code is available at \href{https://github.com/xingmingyu123456/KnowledgeSwapping}{https://github.com/xingmingyu123456/KnowledgeSwapping}.

cs.CV↗

arXiv2025

When Video Coding Meets Multimodal Large Language Models: A Unified Paradigm for Video Coding

Existing codecs are designed to eliminate intrinsic redundancies to create a compact representation for compression. However, strong external priors from Multimodal Large Language Models (MLLMs) have not been explicitly explored in video compression. Herein, we introduce a unified paradigm for Cross-Modality Video Coding (CMVC), which is a pioneering approach to explore multimodality representation and video generative models in video coding. Specifically, on the encoder side, we disentangle a video into spatial content and motion components, which are subsequently transformed into distinct modalities to achieve very compact representation by leveraging MLLMs. During decoding, previously encoded components and video generation models are leveraged to create multiple encoding-decoding modes that optimize video reconstruction quality for specific decoding requirements, including Text-Text-to-Video (TT2V) mode to ensure high-quality semantic information and Image-Text-to-Video (IT2V) mode to achieve superb perceptual consistency. In addition, we propose an efficient frame interpolation model for IT2V mode via Low-Rank Adaption (LoRA) tuning to guarantee perceptual quality, which allows the generated motion cues to behave smoothly. Experiments on benchmarks indicate that TT2V achieves effective semantic reconstruction, while IT2V exhibits competitive perceptual consistency. These results highlight potential directions for future research in video coding.

cs.CV↗

arXiv2025

RaSeRec: Retrieval-Augmented Sequential Recommendation

Although prevailing supervised and self-supervised learning augmented sequential recommendation (SeRec) models have achieved improved performance with powerful neural network architectures, we argue that they still suffer from two limitations: (1) Preference Drift, where models trained on past data can hardly accommodate evolving user preference; and (2) Implicit Memory, where head patterns dominate parametric learning, making it harder to recall long tails. In this work, we explore retrieval augmentation in SeRec, to address these limitations. Specifically, we propose a Retrieval-Augmented Sequential Recommendation framework, named RaSeRec, the main idea of which is to maintain a dynamic memory bank to accommodate preference drifts and retrieve relevant memories to augment user modeling explicitly. It consists of two stages: (i) collaborative-based pre-training, which learns to recommend and retrieve; (ii) retrieval-augmented fine-tuning, which learns to leverage retrieved memories. Extensive experiments on three datasets fully demonstrate the superiority and effectiveness of RaSeRec. The implementation code is available at https://github.com/HITsz-TMG/RaSeRec.

cs.IR↗

arXiv2025

TASAR: Transfer-based Attack on Skeletal Action Recognition

Skeletal sequence data, as a widely employed representation of human actions, are crucial in Human Activity Recognition (HAR). Recently, adversarial attacks have been proposed in this area, which exposes potential security concerns, and more importantly provides a good tool for model robustness test. Within this research, transfer-based attack is an important tool as it mimics the real-world scenario where an attacker has no knowledge of the target model, but is under-explored in Skeleton-based HAR (S-HAR). Consequently, existing S-HAR attacks exhibit weak adversarial transferability and the reason remains largely unknown. In this paper, we investigate this phenomenon via the characterization of the loss function. We find that one prominent indicator of poor transferability is the low smoothness of the loss function. Led by this observation, we improve the transferability by properly smoothening the loss when computing the adversarial examples. This leads to the first Transfer-based Attack on Skeletal Action Recognition, TASAR. TASAR explores the smoothened model posterior of pre-trained surrogates, which is achieved by a new post-train Dual Bayesian optimization strategy. Furthermore, unlike existing transfer-based methods which overlook the temporal coherence within sequences, TASAR incorporates motion dynamics into the Bayesian attack, effectively disrupting the spatial-temporal coherence of S-HARs. For exhaustive evaluation, we build the first large-scale robust S-HAR benchmark, comprising 7 S-HAR models, 10 attack methods, 3 S-HAR datasets and 2 defense models. Extensive results demonstrate the superiority of TASAR. Our benchmark enables easy comparisons for future studies, with the code available in the https://github.com/yunfengdiao/Skeleton-Robustness-Benchmark.

cs.CV↗

arXiv2025

Position reconstruction and surface background model for the PandaX-4T detector

We report the position reconstruction methods and surface background model for the PandaX-4T dark matter direct search experiment. This work develops two position reconstruction algorithms: template matching (TM) method and photon acceptance function (PAF) method. Both methods determine the horizontal position of events based on the light pattern of secondary scintillation collected by the light sensors. After a comprehensive evaluation of resolution, uniformity, and robustness, the PAF method was selected for position reconstruction, while the TM method was employed for verification. The PAF method achieves a bulk event resolution of 1.0 mm and a surface event resolution of 4.4 mm for a typical $S2$ signal with a bottom charge of 1500 PE (about 14 keV). The uniformity is around 20\%. Robustness studies reveal average deviations of 5.1 mm and 8.8 mm for the commissioning run (Run0) and the first science run (Run1), respectively, due to the deactivation of certain PMTs. A data-driven surface background model is developed based on the PAF method. The surface background is estimated to be $0.09 \pm 0.06$ events for Run0 (0.54 tonne$\cdot$year) and $0.17 \pm 0.11$ events for Run1 (1.00 tonne$\cdot$year).

physics.ins-det↗

arXiv2025

Dark Matter Search Results from 1.54 Tonne$\cdot$Year Exposure of PandaX-4T

In this letter, we report the dark matter search results from the commissioning run and the first science run of the PandaX-4T experiment. A blind analysis is carried out on the entire data set. The data processing is improved compared to previous work, unifying the low-level signal reconstruction in a wide energy range up to 120 keV. With a total exposure of 1.54 tonne$\cdot$year, no significant excess of nuclear recoil events is found. The lowest 90% confidence level exclusion on the spin-independent cross section is $1.6 \times 10^{-47}\,\mathrm{cm}^2$ at a dark matter mass of 40 GeV$/c^2$. Our results represent the most stringent constraint for a dark matter mass above 100 GeV$/c^2$.

hep-ex↗

arXiv2025

Exploring New Physics with PandaX-4T Low Energy Electronic Recoil Data

New particles beyond the standard model of particle physics, such as axions, can be effectively searched through their interactions with electrons. We use the large liquid xenon detector PandaX-4T to search for novel electronic recoil signals induced by solar axions, neutrinos with anomalous magnetic moment, axionlike particles, dark photons, and light fermionic dark matter. A detailed background model is established using the latest datasets with 1.54 $\rm ton \cdot yr$ exposure. No significant excess above the background has been observed, and we have obtained competitive constraints for axion couplings, neutrino magnetic moment, and fermionic dark matter interactions.

hep-ex↗

arXiv2025

MOL-Mamba: Enhancing Molecular Representation with Structural & Electronic Insights

Molecular representation learning plays a crucial role in various downstream tasks, such as molecular property prediction and drug design. To accurately represent molecules, Graph Neural Networks (GNNs) and Graph Transformers (GTs) have shown potential in the realm of self-supervised pretraining. However, existing approaches often overlook the relationship between molecular structure and electronic information, as well as the internal semantic reasoning within molecules. This omission of fundamental chemical knowledge in graph semantics leads to incomplete molecular representations, missing the integration of structural and electronic data. To address these issues, we introduce MOL-Mamba, a framework that enhances molecular representation by combining structural and electronic insights. MOL-Mamba consists of an Atom & Fragment Mamba-Graph (MG) for hierarchical structural reasoning and a Mamba-Transformer (MT) fuser for integrating molecular structure and electronic correlation learning. Additionally, we propose a Structural Distribution Collaborative Training and E-semantic Fusion Training framework to further enhance molecular representation learning. Extensive experiments demonstrate that MOL-Mamba outperforms state-of-the-art baselines across eleven chemical-biological molecular datasets.

cs.LG↗

arXiv2025

Towards Pixel-Level Prediction for Gaze Following: Benchmark and Approach

Following the gaze of other people and analyzing the target they are looking at can help us understand what they are thinking, and doing, and predict the actions that may follow. Existing methods for gaze following struggle to perform well in natural scenes with diverse objects, and focus on gaze points rather than objects, making it difficult to deliver clear semantics and accurate scope of the targets. To address this shortcoming, we propose a novel gaze target prediction solution named GazeSeg, that can fully utilize the spatial visual field of the person as guiding information and lead to a progressively coarse-to-fine gaze target segmentation and recognition process. Specifically, a prompt-based visual foundation model serves as the encoder, working in conjunction with three distinct decoding modules (e.g. FoV perception, heatmap generation, and segmentation) to form the framework for gaze target prediction. Then, with the head bounding box performed as an initial prompt, GazeSeg obtains the FoV map, heatmap, and segmentation map progressively, leading to a unified framework for multiple tasks (e.g. direction estimation, gaze target segmentation, and recognition). In particular, to facilitate this research, we construct and release a new dataset, comprising 72k images with pixel-level annotations and 270 categories of gaze targets, built upon the GazeFollow dataset. The quantitative evaluation shows that our approach achieves the Dice of 0.325 in gaze target segmentation and 71.7% top-5 recognition. Meanwhile, our approach also outperforms previous state-of-the-art methods, achieving 0.953 in AUC on the gaze-following task. The dataset and code will be released.

cs.CV↗

arXiv2025

Unsupervised Domain Adaptation via Style-Aware Self-intermediate Domain

Unsupervised domain adaptation (UDA) has attracted considerable attention, which transfers knowledge from a label-rich source domain to a related but unlabeled target domain. Reducing inter-domain differences has always been a crucial factor to improve performance in UDA, especially for tasks where there is a large gap between source and target domains. To this end, we propose a novel style-aware feature fusion method (SAFF) to bridge the large domain gap and transfer knowledge while alleviating the loss of class-discriminative information. Inspired by the human transitive inference and learning ability, a novel style-aware self-intermediate domain (SSID) is investigated to link two seemingly unrelated concepts through a series of intermediate auxiliary synthesized concepts. Specifically, we propose a novel learning strategy of SSID, which selects samples from both source and target domains as anchors, and then randomly fuses the object and style features of these anchors to generate labeled and style-rich intermediate auxiliary features for knowledge transfer. Moreover, we design an external memory bank to store and update specified labeled features to obtain stable class features and class-wise style features. Based on the proposed memory bank, the intra- and inter-domain loss functions are designed to improve the class recognition ability and feature compatibility, respectively. Meanwhile, we simulate the rich latent feature space of SSID by infinite sampling and the convergence of the loss function by mathematical theory. Finally, we conduct comprehensive experiments on commonly used domain adaptive benchmarks to evaluate the proposed SAFF, and the experimental results show that the proposed SAFF can be easily combined with different backbone networks and obtain better performance as a plug-in-plug-out module.

cs.CV↗

arXiv2025

Identification of the superconductivity in bilayer nickelate La$_3$Ni$_2$O$_7$ upon 100 GPa

Identification of superconductivity in the Ruddlesden-Popper phases of nickelates under high pressure remains challenging. Here, we report a comprehensive study of the crystal structure, resistance, and Meissner effect in single crystals of La$_3$Ni$_2$O$_7$ with hydrostatic pressures up to 104 GPa. X-ray diffraction measurements reveal a structural transition from the orthorhombic to a tetragonal phase above 40 GPa. Zero resistance of the superconductivity was achieved with a maximum onset $T_c^{onset}$ of 83 K at 18.0 GPa. Superconductivity is gradually suppressed until it disappears above 80 GPa, resulting in a right-triangle-like superconducting region. The direct-current magnetic susceptibility technique successfully detected the Meissner effect in La$_3$Ni$_2$O$_7$ under pressure; the maximum superconducting volume fraction is estimated to be 62.7% at 22.0 GPa. Thus, we demonstrate the bulk nature of superconductivity in the bilayer nickelate La$_3$Ni$_2$O$_7$ single crystals under high pressure. The results reveal intimate connections among the superconductivity, oxygen content, and structure in La$_3$Ni$_2$O$_7$.

cond-mat.supr-con↗

arXiv2025

Low volume fraction of high-Tc superconductivity in La3Ni2O7 at 80 K and ambient pressure

The discovery of superconductivity in pressurized La3Ni2O7 with a transition temperature of approximately 80 K above the boiling point of liquid nitrogen has sparked significant attention. It is essential to search for high-temperature superconductivity in bulk samples and at ambient pressure in nickelates. In this study, we report influential factors that affect the appearance of superconductivity in La3Ni2O7 at ambient pressure. From direct-current magnetic measurements, we observe a clear diamagnetic response at 80 K in post-annealed single crystals of La3Ni2O7 in oxygen. The superconducting volume fraction is estimated to be within 0.2%, resulting in a decrease in resistivity. This work presents a practical approach for further investigating high-temperature superconductivity in nickelates at ambient pressure.

cond-mat.supr-con↗

arXiv2025

Exploring Heterogeneity and Uncertainty for Graph-based Cognitive Diagnosis Models in Intelligent Education

Graph-based Cognitive Diagnosis (CD) has attracted much research interest due to its strong ability on inferring students' proficiency levels on knowledge concepts. While graph-based CD models have demonstrated remarkable performance, we contend that they still cannot achieve optimal performance due to the neglect of edge heterogeneity and uncertainty. Edges involve both correct and incorrect response logs, indicating heterogeneity. Meanwhile, a response log can have uncertain semantic meanings, e.g., a correct log can indicate true mastery or fortunate guessing, and a wrong log can indicate a lack of understanding or a careless mistake. In this paper, we propose an Informative Semantic-aware Graph-based Cognitive Diagnosis model (ISG-CD), which focuses on how to utilize the heterogeneous graph in CD and minimize effects of uncertain edges. Specifically, to explore heterogeneity, we propose a semantic-aware graph neural networks based CD model. To minimize effects of edge uncertainty, we propose an Informative Edge Differentiation layer from an information bottleneck perspective, which suggests keeping a minimal yet sufficient reliable graph for CD in an unsupervised way. We formulate this process as maximizing mutual information between the reliable graph and response logs, while minimizing mutual information between the reliable graph and the original graph. After that, we prove that mutual information maximization can be theoretically converted to the classic binary cross entropy loss function, while minimizing mutual information can be realized by the Hilbert-Schmidt Independence Criterion. Finally, we adopt an alternating training strategy for optimizing learnable parameters of both the semantic-aware graph neural networks based CD model and the edge differentiation layer. Extensive experiments on three real-world datasets have demonstrated the effectiveness of ISG-CD.

cs.CY↗

arXiv2025

Less is More: Information Bottleneck Denoised Multimedia Recommendation

Empowered by semantic-rich content information, multimedia recommendation has emerged as a potent personalized technique. Current endeavors center around harnessing multimedia content to refine item representation or uncovering latent item-item structures based on modality similarity. Despite the effectiveness, we posit that these methods are usually suboptimal due to the introduction of irrelevant multimedia features into recommendation tasks. This stems from the fact that generic multimedia feature extractors, while well-designed for domain-specific tasks, can inadvertently introduce task-irrelevant features, leading to potential misguidance of recommenders. In this work, we propose a denoised multimedia recommendation paradigm via the Information Bottleneck principle (IB). Specifically, we propose a novel Information Bottleneck denoised Multimedia Recommendation (IBMRec) model to tackle the irrelevant feature issue. IBMRec removes task-irrelevant features from both feature and item-item structure perspectives, which are implemented by two-level IB learning modules: feature-level (FIB) and graph-level (GIB). In particular, FIB focuses on learning the minimal yet sufficient multimedia features. This is achieved by maximizing the mutual information between multimedia representation and recommendation tasks, while concurrently minimizing it between multimedia representation and pre-trained multimedia features. Furthermore, GIB is designed to learn the robust item-item graph structure, it refines the item-item graph based on preference affinity, then minimizes the mutual information between the original graph and the refined one. Extensive experiments across three benchmarks validate the effectiveness of our proposed model, showcasing high performance, and applicability to various multimedia recommenders.

cs.IR↗

arXiv2025

Unveiling an in-plane Hall effect in rutile RuO$_2$ films

The in-plane-magnetic-field-induced Hall effect (IPHE) observed in Weyl semimetals and PT-symmetric antiferromagnets has attracted increasing attention, as it breaks the stereotype that the Hall effect is induced by an out-of-plane magnetic field or magnetization. To date, the IPHE has been discussed mainly for materials with low-symmetry crystal/magnetic point groups. Here, we show that even if symmetry forbids an inherent IPHE that arises from any mechanism, an apparent IPHE can be generated by selecting a low-symmetry crystalline plane for measurement. For rutile RuO$_2$, although its high symmetry forbids an inherent IPHE, films grown along the low-symmetry (1 1 1) and (1 0 1) orientations are found to exhibit a distinct IPHE. The in-plane Hall coefficients are quantitatively reproduced by referring to the out-of-plane Hall coefficients measured for the high-symmetry (1 0 0) and (0 0 1) planes, indicating that the observed IPHE is caused by a superposition of inequivalent out-of-plane Hall effects. Similar behaviour is also observed for paramagnetic rutile systems, indicating the ubiquity of the apparent IPHE in electronic and spintronic devices with low-symmetry crystalline planes.

cond-mat.mtrl-sci↗

arXiv2025

Towards Lightweight Time Series Forecasting: a Patch-wise Transformer with Weak Data Enriching

Patch-wise Transformer based time series forecasting achieves superior accuracy. However, this superiority relies heavily on intricate model design with massive parameters, rendering both training and inference expensive, thus preventing their deployments on edge devices with limited resources and low latency requirements. In addition, existing methods often work in an autoregressive manner, which take into account only historical values, but ignore valuable, easy-to-obtain context information, such as weather forecasts, date and time of day. To contend with the two limitations, we propose LiPFormer, a novel Lightweight Patch-wise Transformer with weak data enriching. First, to simplify the Transformer backbone, LiPFormer employs a novel lightweight cross-patch attention and a linear transformation-based attention to eliminate Layer Normalization and Feed Forward Network, two heavy components in existing Transformers. Second, we propose a lightweight, weak data enriching module to provide additional, valuable weak supervision to the training. It enhances forecasting accuracy without significantly increasing model complexity as it does not involve expensive, human-labeling but using easily accessible context information. This facilitates the weak data enriching to plug-and-play on existing models. Extensive experiments on nine benchmark time series datasets demonstrate that LiPFormer outperforms state-of-the-art methods in accuracy, while significantly reducing parameter scale, training duration, and GPU memory usage. Deployment on an edge device reveals that LiPFormer takes only 1/3 inference time compared to classic Transformers. In addition, we demonstrate that the weak data enriching can integrate seamlessly into various Transformer based models to enhance their accuracy, suggesting its generality.

cs.LG↗

arXiv2025

Exploring Power Side-Channel Challenges in Embedded Systems Security

Power side-channel (PSC) attacks are widely used in embedded microcontrollers, particularly in cryptographic applications, to extract sensitive information. However, expanding the applications of PSC attacks to broader security contexts in the embedded systems domain faces significant challenges. These include the need for specialized hardware setups to manage high noise levels in real-world targets and assumptions regarding the attacker's knowledge and capabilities. This paper systematically analyzes these challenges and introduces a novel signal-processing method that addresses key limitations, enabling effective PSC attacks in real-world embedded systems without requiring hardware modifications. We validate the proposed approach through experiments on real-world black-box embedded devices, verifying its potential to expand its usage in various embedded systems security applications beyond traditional cryptographic applications.

cs.CR↗

arXiv2025

Structure of an axisymmetric turbulent boundary layer under adverse pressure gradient: a large-eddy simulation study

The spatial characteristics and structure of an axisymmetric turbulent boundary layer under strong adverse pressure gradient and weak transverse curvature are investigated using incompressible large-eddy simulation. The boundary layer is on a $20^{\circ}$ tail cone of a body of revolution at a length-based Reynolds number of $1.9\times10^6$. The simulation results are in agreement with the experimental measurements of Balantrapu et al. (J. Fluid Mech., vol. 929, 2021) and significantly expand the experimental results with new flow-field details and physical insights. The mean streamwise velocity profiles exhibit a shortened logarithmic region and a longer wake region compared with planar boundary layers at zero pressure gradient. With the embedded-shear-layer scaling, self-similarity is observed for the mean velocity and all three components of turbulence intensity. The azimuthal-wavenumber spectra of streamwise velocity fluctuations possess two peaks in the wall-normal direction, an inner peak at the wavelength of approximately 100 wall units and an outer peak in the wake region with growing strength, wavelength and distance to the wall in the downstream direction. Two-point correlations of streamwise velocity fluctuations show significant downstream growth and elongation of turbulence structures with increasing inclination angle. However, relative to the boundary-layer thickness, the correlation structures decrease in size in the downstream direction. The distributions of streamwise and wall-normal integral lengths across the boundary-layer thickness resemble those of zero-pressure-gradient planar boundary layers, whereas the azimuthal integral length deviates from the planar boundary-layer behavior at downstream stations.

physics.flu-dyn↗