arXiv subjects

Hui Ma

Publications and source records attributed to Hui Ma.

At least 73 records · Page 4Linked to original sources

arXiv2025

Efficient Tuning of Large Language Models for Knowledge-Grounded Dialogue Generation

Large language models (LLMs) demonstrate remarkable text comprehension and generation capabilities but often lack the ability to utilize up-to-date or domain-specific knowledge not included in their training data. To address this gap, we introduce KEDiT, an efficient method for fine-tuning LLMs for knowledge-grounded dialogue generation. KEDiT operates in two main phases: first, it employs an information bottleneck to compress retrieved knowledge into learnable parameters, retaining essential information while minimizing computational overhead. Second, a lightweight knowledge-aware adapter integrates these compressed knowledge vectors into the LLM during fine-tuning, updating less than 2\% of the model parameters. The experimental results on the Wizard of Wikipedia and a newly constructed PubMed-Dialog dataset demonstrate that KEDiT excels in generating contextually relevant and informative responses, outperforming competitive baselines in automatic, LLM-based, and human evaluations. This approach effectively combines the strengths of pretrained LLMs with the adaptability needed for incorporating dynamic knowledge, presenting a scalable solution for fields such as medicine.

cs.CL↗

arXiv2025

Classification results for totally real surfaces of nearly Kähler $\mathbb{C}P^3$

Totally real surfaces in the nearly Kähler $\mathbb{C}P^3$ are investigated and are completely classified under various additional assumptions, resulting in multiple new examples. Among others, the classification includes totally real surfaces that are extrinsically homogeneous; or minimal; or totally umbilical; or Codazzi-like (including parallel and non-parallel examples).

math.DG↗

arXiv2025

Domain Generalization for Face Anti-spoofing via Content-aware Composite Prompt Engineering

The challenge of Domain Generalization (DG) in Face Anti-Spoofing (FAS) is the significant interference of domain-specific signals on subtle spoofing clues. Recently, some CLIP-based algorithms have been developed to alleviate this interference by adjusting the weights of visual classifiers. However, our analysis of this class-wise prompt engineering suffers from two shortcomings for DG FAS: (1) The categories of facial categories, such as real or spoof, have no semantics for the CLIP model, making it difficult to learn accurate category descriptions. (2) A single form of prompt cannot portray the various types of spoofing. In this work, instead of class-wise prompts, we propose a novel Content-aware Composite Prompt Engineering (CCPE) that generates instance-wise composite prompts, including both fixed template and learnable prompts. Specifically, our CCPE constructs content-aware prompts from two branches: (1) Inherent content prompt explicitly benefits from abundant transferred knowledge from the instruction-based Large Language Model (LLM). (2) Learnable content prompts implicitly extract the most informative visual content via Q-Former. Moreover, we design a Cross-Modal Guidance Module (CGM) that dynamically adjusts unimodal features for fusion to achieve better generalized FAS. Finally, our CCPE has been validated for its effectiveness in multiple cross-domain experiments and achieves state-of-the-art (SOTA) results.

cs.CV↗

arXiv2025

FA^{3}-CLIP: Frequency-Aware Cues Fusion and Attack-Agnostic Prompt Learning for Unified Face Attack Detection

Facial recognition systems are vulnerable to physical (e.g., printed photos) and digital (e.g., DeepFake) face attacks. Existing methods struggle to simultaneously detect physical and digital attacks due to: 1) significant intra-class variations between these attack types, and 2) the inadequacy of spatial information alone to comprehensively capture live and fake cues. To address these issues, we propose a unified attack detection model termed Frequency-Aware and Attack-Agnostic CLIP (FA\textsuperscript{3}-CLIP), which introduces attack-agnostic prompt learning to express generic live and fake cues derived from the fusion of spatial and frequency features, enabling unified detection of live faces and all categories of attacks. Specifically, the attack-agnostic prompt module generates generic live and fake prompts within the language branch to extract corresponding generic representations from both live and fake faces, guiding the model to learn a unified feature space for unified attack detection. Meanwhile, the module adaptively generates the live/fake conditional bias from the original spatial and frequency information to optimize the generic prompts accordingly, reducing the impact of intra-class variations. We further propose a dual-stream cues fusion framework in the vision branch, which leverages frequency information to complement subtle cues that are difficult to capture in the spatial domain. In addition, a frequency compression block is utilized in the frequency stream, which reduces redundancy in frequency features while preserving the diversity of crucial cues. We also establish new challenging protocols to facilitate unified face attack detection effectiveness. Experimental results demonstrate that the proposed method significantly improves performance in detecting physical and digital face attacks, achieving state-of-the-art results.

cs.CV↗

arXiv2025

Empathy Level Alignment via Reinforcement Learning for Empathetic Response Generation

Empathetic response generation, aiming to understand the user's situation and feelings and respond empathically, is crucial in building human-like dialogue systems. Traditional approaches typically employ maximum likelihood estimation as the optimization objective during training, yet fail to align the empathy levels between generated and target responses. To this end, we propose an empathetic response generation framework using reinforcement learning (EmpRL). The framework develops an effective empathy reward function and generates empathetic responses by maximizing the expected reward through reinforcement learning. EmpRL utilizes the pre-trained T5 model as the generator and further fine-tunes it to initialize the policy. To align the empathy levels between generated and target responses within a given context, an empathy reward function containing three empathy communication mechanisms -- emotional reaction, interpretation, and exploration -- is constructed using pre-designed and pre-trained empathy identifiers. During reinforcement learning training, the proximal policy optimization algorithm is used to fine-tune the policy, enabling the generation of empathetic responses. Both automatic and human evaluations demonstrate that the proposed EmpRL framework significantly improves the quality of generated responses, enhances the similarity in empathy levels between generated and target responses, and produces empathetic responses covering both affective and cognitive aspects.

cs.CL↗

arXiv2025

Distilling Implicit Multimodal Knowledge into Large Language Models for Zero-Resource Dialogue Generation

Integrating multimodal knowledge into large language models (LLMs) represents a significant advancement in dialogue generation capabilities. However, the effective incorporation of such knowledge in zero-resource scenarios remains a substantial challenge due to the scarcity of diverse, high-quality dialogue datasets. To address this, we propose the Visual Implicit Knowledge Distillation Framework (VIKDF), an innovative approach aimed at enhancing LLMs for enriched dialogue generation in zero-resource contexts by leveraging implicit multimodal knowledge. VIKDF comprises two main stages: knowledge distillation, using an Implicit Query Transformer to extract and encode visual implicit knowledge from image-text pairs into knowledge vectors; and knowledge integration, employing a novel Bidirectional Variational Information Fusion technique to seamlessly integrate these distilled vectors into LLMs. This enables the LLMs to generate dialogues that are not only coherent and engaging but also exhibit a deep understanding of the context through implicit multimodal cues, effectively overcoming the limitations of zero-resource scenarios. Our extensive experimentation across two dialogue datasets shows that VIKDF outperforms existing state-of-the-art models in generating high-quality dialogues. The code is available at https://github.com/zhangbo-nlp/VIKDF.

cs.CL↗

arXiv2025

Unveiling potential candidates for rare-earth-free permanent magnet and magnetocaloric effect applications: a high throughput screening in Fe-N alloys

Based on high-throughput density functional theory calculations, we have found 49 ferromag-netic cases in FexN1-x (0<x<1) compounds, focusing especially on permanent magnet and giant magnetocaloric effect applications. It is found that 15 compounds are potential permanent mag-nets with a magneto-crystalline anisotropy energy more than 1 MJ/m3, filling in the gap of appli-cation spectrum between high-performance and widely used permanents. Among the potential permanent magnets, Fe2N can be classified as a hard magnet while the other 14 compounds can be classified as semi-hard magnets. According to the calculations of magnetic deformation proxy, 40 compounds are identified as potential giant magnetocaloric effect candidates. We suspect that Fe-N compounds provide fine opportunities for applications in both rare-earth free permanent magnets and magnetocaloric effect.

cond-mat.mtrl-sci↗

arXiv2024

Cost-Aware Dynamic Cloud Workflow Scheduling using Self-Attention and Evolutionary Reinforcement Learning

The Cost-aware Dynamic Multi-Workflow Scheduling (CDMWS) in the cloud is a kind of cloud workflow management problem, which aims to assign virtual machine (VM) instances to execute tasks in workflows so as to minimize the total costs, including both the penalties for violating Service Level Agreement (SLA) and the VM rental fees. Powered by deep neural networks, Reinforcement Learning (RL) methods can construct effective scheduling policies for solving CDMWS problems. Traditional policy networks in RL often use basic feedforward architectures to separately determine the suitability of assigning any VM instances, without considering all VMs simultaneously to learn their global information. This paper proposes a novel self-attention policy network for cloud workflow scheduling (SPN-CWS) that captures global information from all VMs. We also develop an Evolution Strategy-based RL (ERL) system to train SPN-CWS reliably and effectively. The trained SPN-CWS can effectively process all candidate VM instances simultaneously to identify the most suitable VM instance to execute every workflow task. Comprehensive experiments show that our method can noticeably outperform several state-of-the-art algorithms on multiple benchmark CDMWS problems.

cs.AI↗

arXiv2024

Anisotropic capillary hypersurfaces in a wedge

We investigate anisotropic capillary hypersurfaces within a wedge in Euclidean space. In this study, we generalize the Minkowski norm $F$, traditionally employed to define the anisotropic surface energy, to a gauge on the unit sphere $S^n$. This generalization helps to illuminate a significant relationship between capillary hypersurfaces and hypersurfaces with free boundary. Our main results include new Minkowski formulae and a Heintze-Karcher type inequality. As an application, we prove an Alexandrov-type theorem, thereby extending the known results to the anisotropic setting.

math.DG↗

arXiv2024

PsycoLLM: Enhancing LLM for Psychological Understanding and Evaluation

Mental health has attracted substantial attention in recent years and LLM can be an effective technology for alleviating this problem owing to its capability in text understanding and dialogue. However, existing research in this domain often suffers from limitations, such as training on datasets lacking crucial prior knowledge and evidence, and the absence of comprehensive evaluation methods. In this paper, we propose a specialized psychological large language model (LLM), named PsycoLLM, trained on a proposed high-quality psychological dataset, including single-turn QA, multi-turn dialogues and knowledge-based QA. Specifically, we construct multi-turn dialogues through a three-step pipeline comprising multi-turn QA generation, evidence judgment, and dialogue refinement. We augment this process with real-world psychological case backgrounds extracted from online platforms, enhancing the relevance and applicability of the generated data. Additionally, to compare the performance of PsycoLLM with other LLMs, we develop a comprehensive psychological benchmark based on authoritative psychological counseling examinations in China, which includes assessments of professional ethics, theoretical proficiency, and case analysis. The experimental results on the benchmark illustrate the effectiveness of PsycoLLM, which demonstrates superior performance compared to other LLMs.

cs.CL↗

arXiv2024

The Fe-N system: crystal structure prediction, phase stability, and mechanical properties

Nitriding introduces nitrides into the surface of steels, significantly enhancing the surface me-chanical properties. By combining the variable composition evolutionary algorithm and first-principles calculations based on density functional theory, 50 thermodynamically stable or metastable Fe-N compounds with various stoichiometric ratios were identified, exhibiting also dynamic and mechanical stability. The mechanical properties of these structures were systemati-cally studied, including the bulk modulus, shear modulus, Young's modulus, Poisson's ratio, Pugh's ratio, Cauchy pressure, Klemen parameters, universal elastic anisotropy, Debye tempera-ture, and Vickers hardness. All identified stable and metastable Fe-N compounds were found in the ductile region, with most exhibiting homogeneous elastic properties and isotropic metallic bonding. As the nitrogen concentration increases, their bulk moduli generally increase as well. The Vickers hardness values of Fe-N compounds range from 3.5 to 10.5 GPa, which are signifi-cantly higher than that of pure Fe (2.0 GPa), due to the stronger Fe-N bonds strength. This study provides insights into optimizing and designing Fe-N alloys with tailored mechanical properties.

cond-mat.mtrl-sci↗

arXiv2024

Uncertainty-aware t-distributed Stochastic Neighbor Embedding for Single-cell RNA-seq Data

Nonlinear data visualization using t-distributed stochastic neighbor embedding (t-SNE) enables the representation of complex single-cell transcriptomic landscapes in two or three dimensions to depict biological populations accurately. However, t-SNE often fails to account for uncertainties in the original dataset, leading to misleading visualizations where cell subsets with noise appear indistinguishable. To address these challenges, we introduce uncertainty-aware t-SNE (Ut-SNE), a noise-defending visualization tool tailored for uncertain single-cell RNA-seq data. By creating a probabilistic representation for each sample, Our Ut-SNE accurately incorporates noise about transcriptomic variability into the visual interpretation of single-cell RNA sequencing data, revealing significant uncertainties in transcriptomic variability. Through various examples, we showcase the practical value of Ut-SNE and underscore the significance of incorporating uncertainty awareness into data visualization practices. This versatile uncertainty-aware visualization tool can be easily adapted to other scientific domains beyond single-cell RNA sequencing, making them valuable resources for high-dimensional data analysis.

q-bio.GN↗

arXiv2024

Machine Learning Model for Complete Reconstruction of Diagnostic Polarimetric Images from partial Mueller polarimetry data

The translation of imaging Mueller polarimetry to clinical practice is often hindered by large footprint and relatively slow acquisition speed of the existing instruments. Using polarization-sensitive camera as a detector may reduce instrument dimensions and allow data streaming at video rate. However, only the first three rows of a complete 4x4 Mueller matrix can be measured. To overcome this hurdle we developed a machine learning approach using sequential neural network algorithm for the reconstruction of missing elements of a Mueller matrix from the measured elements of the first three rows. The algorithm was trained and tested on the dataset of polarimetric images of various excised human tissues (uterine cervix, colon, skin, brain) acquired with two different imaging Mueller polarimeters operating in either reflection (wide-field imaging system) or transmission (microscope) configurations at different wavelengths of 550 nm and 385 nm, respectively. The reconstruction performance was evaluated using various error metrics, all of which confirmed low error values. The execution time of the trained neural network algorithm was about 300 microseconds for a single image pixel. It suggests that a machine learning approach with parallel processing of all image pixels combined with the partial Mueller polarimeter operating at video rate can effectively substitute for the complete Mueller polarimeter and produce accurate maps of depolarization, linear retardance and orientation of the optical axis of biological tissues, which can be used for medical diagnosis in clinical settings.

physics.optics↗

arXiv2024

Online t-SNE for single-cell RNA-seq

Due to the sequential sample arrival, changing experiment conditions, and evolution of knowledge, the demand to continually visualize evolving structures of sequential and diverse single-cell RNA-sequencing (scRNA-seq) data becomes indispensable. However, as one of the state-of-the-art visualization and analysis methods for scRNA-seq, t-distributed stochastic neighbor embedding (t-SNE) merely visualizes static scRNA-seq data offline and fails to meet the demand well. To address these challenges, we introduce online t-SNE to seamlessly integrate sequential scRNA-seq data. Online t-SNE achieves this by leveraging the embedding space of old samples, exploring the embedding space of new samples, and aligning the two embedding spaces on the fly. Consequently, online t-SNE dramatically enables the continual discovery of new structures and high-quality visualization of new scRNA-seq data without retraining from scratch. We showcase the formidable visualization capabilities of online t-SNE across diverse sequential scRNA-seq datasets.

q-bio.GN↗

arXiv2024

On-Chip Vectorial Structured Light Manipulation via Inverse Design

On-chip structured light, with potentially infinite complexity, has emerged as a linchpin in the realm of integrated photonics. However, the realization of arbitrarily tailoring a multitude of light field dimensions in complex media remains a challenge1, Through associating physical light fields and mathematical function spaces by introducing a mapping operator, we proposed a data-driven inverse design method to precisely manipulate between any two structured light fields in the on-chip high-dimensional Hilbert space. To illustrate, light field conversion in on-chip topological photonics was achieved. High-performance topological coupling devices with minimal insertion loss and customizable topological routing devices were designed and realized. Our method provides a new paradigm to enable precise manipulation over the on-chip vectorial structured light and paves the way for the realization of complex photonic functions.

physics.optics↗

arXiv2024

FER-YOLO-Mamba: Facial Expression Detection and Classification Based on Selective State Space

Facial Expression Recognition (FER) plays a pivotal role in understanding human emotional cues. However, traditional FER methods based on visual information have some limitations, such as preprocessing, feature extraction, and multi-stage classification procedures. These not only increase computational complexity but also require a significant amount of computing resources. Considering Convolutional Neural Network (CNN)-based FER schemes frequently prove inadequate in identifying the deep, long-distance dependencies embedded within facial expression images, and the Transformer's inherent quadratic computational complexity, this paper presents the FER-YOLO-Mamba model, which integrates the principles of Mamba and YOLO technologies to facilitate efficient coordination in facial expression image recognition and localization. Within the FER-YOLO-Mamba model, we further devise a FER-YOLO-VSS dual-branch module, which combines the inherent strengths of convolutional layers in local feature extraction with the exceptional capability of State Space Models (SSMs) in revealing long-distance dependencies. To the best of our knowledge, this is the first Vision Mamba model designed for facial expression detection and classification. To evaluate the performance of the proposed FER-YOLO-Mamba model, we conducted experiments on two benchmark datasets, RAF-DB and SFEW. The experimental results indicate that the FER-YOLO-Mamba model achieved better results compared to other models. The code is available from https://github.com/SwjtuMa/FER-YOLO-Mamba.

cs.CV↗

arXiv2024

Minimal Lagrangian surfaces in $\mathbb{C}P^2$ via the loop group method Part II: The general case

We extend the techniques introduced in \cite{DoMaB1} for contractible Riemann surfaces to construct minimal Lagrangian immersions from arbitrary Riemann surfaces into $\mathbb{C}P^2$ via the loop group method. Based on the potentials of translationally equivariant minimal Lagrangian surfaces, we introduce perturbed equivariant minimal Lagrangian surfaces in $\mathbb{C}P^2$ and construct a class of minimal Lagrangian cylinders. Furthermore, we show that these minimal Lagrangian cylinders approximate Delaunay cylinders with respect to some weighted Wiener norm of the twisted loop group $ΛSU(3)_σ$.

math.DG↗

arXiv2024

Joint Physical-Digital Facial Attack Detection Via Simulating Spoofing Clues

Face recognition systems are frequently subjected to a variety of physical and digital attacks of different types. Previous methods have achieved satisfactory performance in scenarios that address physical attacks and digital attacks, respectively. However, few methods are considered to integrate a model that simultaneously addresses both physical and digital attacks, implying the necessity to develop and maintain multiple models. To jointly detect physical and digital attacks within a single model, we propose an innovative approach that can adapt to any network architecture. Our approach mainly contains two types of data augmentation, which we call Simulated Physical Spoofing Clues augmentation (SPSC) and Simulated Digital Spoofing Clues augmentation (SDSC). SPSC and SDSC augment live samples into simulated attack samples by simulating spoofing clues of physical and digital attacks, respectively, which significantly improve the capability of the model to detect "unseen" attack types. Extensive experiments show that SPSC and SDSC can achieve state-of-the-art generalization in Protocols 2.1 and 2.2 of the UniAttackData dataset, respectively. Our method won first place in "Unified Physical-Digital Face Attack Detection" of the 5th Face Anti-spoofing Challenge@CVPR2024. Our final submission obtains 3.75% APCER, 0.93% BPCER, and 2.34% ACER, respectively. Our code is available at https://github.com/Xianhua-He/cvpr2024-face-anti-spoofing-challenge.

cs.CV↗