arXiv subjects

Feng Zhou

Publications and source records attributed to Feng Zhou.

At least 73 records · Page 4Linked to original sources

arXiv2025

The Mediating Effects of Emotions on Trust through Risk Perception and System Performance in Automated Driving

Trust in automated vehicles (AVs) has traditionally been explored through a cognitive lens, but growing evidence highlights the significant role emotions play in shaping trust. This study investigates how risk perception and AV performance (error vs. no error) influence emotional responses and trust in AVs, using mediation analysis to examine the indirect effects of emotions. In this study, 70 participants (42 male, 28 female) watched real-life recorded videos of AVs operating with or without errors, coupled with varying levels of risk information (high, low, or none). They reported their anticipated emotional responses using 19 discrete emotion items, and trust was assessed through dispositional, learned, and situational trust measures. Factor analysis identified four key emotional components, namely hostility, confidence, anxiety, and loneliness, that were influenced by risk perception and AV performance. The linear mixed model showed that risk perception was not a significant predictor of trust, while performance and individual differences were. Mediation analysis revealed that confidence was a strong positive mediator, while hostile and anxious emotions negatively impacted trust. However, lonely emotions did not significantly mediate the relationship between AV performance and trust. The results show that real-time AV behavior is more influential on trust than pre-existing risk perceptions, indicating trust in AVs might be more experience-based than shaped by prior beliefs. Our findings also underscore the importance of fostering positive emotional responses for trust calibration, which has important implications for user experience design in automated driving.

cs.HC↗

arXiv2025

Human Motion Instruction Tuning

This paper presents LLaMo (Large Language and Human Motion Assistant), a multimodal framework for human motion instruction tuning. In contrast to conventional instruction-tuning approaches that convert non-linguistic inputs, such as video or motion sequences, into language tokens, LLaMo retains motion in its native form for instruction tuning. This method preserves motion-specific details that are often diminished in tokenization, thereby improving the model's ability to interpret complex human behaviors. By processing both video and motion data alongside textual inputs, LLaMo enables a flexible, human-centric analysis. Experimental evaluations across high-complexity domains, including human behaviors and professional activities, indicate that LLaMo effectively captures domain-specific knowledge, enhancing comprehension and prediction in motion-intensive scenarios. We hope LLaMo offers a foundation for future multimodal AI systems with broad applications, from sports analytics to behavioral prediction. Our code and models are available on the project website: https://github.com/ILGLJ/LLaMo.

cs.AI↗

arXiv2025

Somatic Safety: An Embodied Approach Towards Safe Human-Robot Interaction

As robots enter the messy human world so the vital matter of safety takes on a fresh complexion with physical contact becoming inevitable and even desirable. We report on an artistic-exploration of how dancers, working as part of a multidisciplinary team, engaged in contact improvisation exercises to explore the opportunities and challenges of dancing with cobots. We reveal how they employed their honed bodily senses and physical skills to engage with the robots aesthetically and yet safely, interleaving improvised physical manipulations with reflections to grow their knowledge of how the robots behaved and felt. We introduce somatic safety, a holistic mind-body approach in which safety is learned, felt and enacted through bodily contact with robots in addition to being reasoned about. We conclude that robots need to be better designed for people to hold them and might recognise tacit safety cues among people.We propose that safety should be learned through iterative bodily experience interleaved with reflection.

cs.HC↗

arXiv2025

Image is All You Need to Empower Large-scale Diffusion Models for In-Domain Generation

In-domain generation aims to perform a variety of tasks within a specific domain, such as unconditional generation, text-to-image, image editing, 3D generation, and more. Early research typically required training specialized generators for each unique task and domain, often relying on fully-labeled data. Motivated by the powerful generative capabilities and broad applications of diffusion models, we are driven to explore leveraging label-free data to empower these models for in-domain generation. Fine-tuning a pre-trained generative model on domain data is an intuitive but challenging way and often requires complex manual hyper-parameter adjustments since the limited diversity of the training data can easily disrupt the model's original generative capabilities. To address this challenge, we propose a guidance-decoupled prior preservation mechanism to achieve high generative quality and controllability by image-only data, inspired by preserving the pre-trained model from a denoising guidance perspective. We decouple domain-related guidance from the conditional guidance used in classifier-free guidance mechanisms to preserve open-world control guidance and unconditional guidance from the pre-trained model. We further propose an efficient domain knowledge learning technique to train an additional text-free UNet copy to predict domain guidance. Besides, we theoretically illustrate a multi-guidance in-domain generation pipeline for a variety of generative tasks, leveraging multiple guidances from distinct diffusion models and conditions. Extensive experiments demonstrate the superiority of our method in domain-specific synthesis and its compatibility with various diffusion-based control methods and applications.

cs.CV↗

arXiv2025

Magnetoelastic coupling in intercalated transition metal dichalcogenides

The large van der Waals gap in transition metal dichalcogenides (TMDs) offers an avenue to host external metal atoms that modify the ground state of these 2D materials. Here, we experimentally and theoretically address the charge correlations in a family of intercalated TMDs. While short-range charge fluctuations develop in Co$_{1/3}$TaS$_{2}$ and Fe$_{1/3}$TaS$_{2}$, long-range charge order switches-on in Fe$_{1/3}$NbS$_{2}$ driven by the interplay of magnetic order and lattice degrees of freedom. The magnetoelastic coupling is demonstrated in Fe$_{1/3}$NbS$_{2}$ by the enhancement of the charge modulations upon magnetic field below T$_\mathrm{N}$, although Density Functional Perturbation Theory (DFPT) calculations predict negligible electron(spin)-phonon coupling. Furthermore, we show that Co-intercalated TaS$_2$ displays a kagome-like Fermi surface, hence opening the path to engineer electronic band structures and study the entanglement of spin, charge, and spin-phonon mechanisms in the large family of intercalated TMDs.

cond-mat.str-el↗

arXiv2025

L2HCount:Generalizing Crowd Counting from Low to High Crowd Density via Density Simulation

Since COVID-19, crowd-counting tasks have gained wide applications. While supervised methods are reliable, annotation is more challenging in high-density scenes due to small head sizes and severe occlusion, whereas it's simpler in low-density scenes. Interestingly, can we train the model in low-density scenes and generalize it to high-density scenes? Therefore, we propose a low- to high-density generalization framework (L2HCount) that learns the pattern related to high-density scenes from low-density ones, enabling it to generalize well to high-density scenes. Specifically, we first introduce a High-Density Simulation Module and a Ground-Truth Generation Module to construct fake high-density images along with their corresponding ground-truth crowd annotations respectively by image-shifting technique, effectively simulating high-density crowd patterns. However, the simulated images have two issues: image blurring and loss of low-density image characteristics. Therefore, we second propose a Head Feature Enhancement Module to extract clear features in the simulated high-density scene. Third, we propose a Dual-Density Memory Encoding Module that uses two crowd memories to learn scene-specific patterns from low- and simulated high-density scenes, respectively. Extensive experiments on four challenging datasets have shown the promising performance of L2HCount.

cs.CV↗

arXiv2025

Exploring Position Encoding in Diffusion U-Net for Training-free High-resolution Image Generation

Denoising higher-resolution latents via a pre-trained U-Net leads to repetitive and disordered image patterns. Although recent studies make efforts to improve generative quality by aligning denoising process across original and higher resolutions, the root cause of suboptimal generation is still lacking exploration. Through comprehensive analysis of position encoding in U-Net, we attribute it to inconsistent position encoding, sourced by the inadequate propagation of position information from zero-padding to latent features in convolution layers as resolution increases. To address this issue, we propose a novel training-free approach, introducing a Progressive Boundary Complement (PBC) method. This method creates dynamic virtual image boundaries inside the feature map to enhance position information propagation, enabling high-quality and rich-content high-resolution image synthesis. Extensive experiments demonstrate the superiority of our method.

cs.CV↗

arXiv2025

Integrating Protein Dynamics into Structure-Based Drug Design via Full-Atom Stochastic Flows

The dynamic nature of proteins, influenced by ligand interactions, is essential for comprehending protein function and progressing drug discovery. Traditional structure-based drug design (SBDD) approaches typically target binding sites with rigid structures, limiting their practical application in drug development. While molecular dynamics simulation can theoretically capture all the biologically relevant conformations, the transition rate is dictated by the intrinsic energy barrier between them, making the sampling process computationally expensive. To overcome the aforementioned challenges, we propose to use generative modeling for SBDD considering conformational changes of protein pockets. We curate a dataset of apo and multiple holo states of protein-ligand complexes, simulated by molecular dynamics, and propose a full-atom flow model (and a stochastic version), named DynamicFlow, that learns to transform apo pockets and noisy ligands into holo pockets and corresponding 3D ligand molecules. Our method uncovers promising ligand molecules and corresponding holo conformations of pockets. Additionally, the resultant holo-like states provide superior inputs for traditional SBDD approaches, playing a significant role in practical drug discovery.

q-bio.BM↗

arXiv2025

Normal conformal metrics with prescribed $Q$-Curvature in $\mathbb{R}^{2n}$

We consider the $Q$-curvature equation \begin{equation}\label{0.1} (-Δ)^n u = K(x)e^{2nu}\quad\text{in} ~\mathbb{R}^{2n} \ (n \geq 2) \end{equation} where $K$ is a given non constant continuous function. Under mild growth control on $K$, we get a necessary condition on the total curvature $Λ_u$ for any normal conformal metric $g_u = e^{2u}|dx|^2$ satisfying $Q_{g_u} = K$ in $\mathbb{R}^{2n}$, or equivalently, solutions to equation with logarithmic growth at infinity. Inversely, when $K$ is nonpositive satisfying polynomial growth control, we show the existence of normal conformal metrics with quasi optimal range of total curvature and precise asymptotic behavior at infinity. If furthermore $K$ is radial symmetric, we establish the same existence result without any growth assumption on $K$.

math.AP↗

arXiv2025

Unconventional topological Weyl-dipole phonon

A pair of Weyl points (WPs) with opposite Chern numbers ${\cal{C}}$ can exhibit an additional higher-order $Z_2$ topological charge, giving rise to the formation of a $Z_2$ Weyl dipole. Owing to the nontrivial topological charge, $Z_2$ Weyl dipoles should also appear in pairs, and the WPs within each $Z_2$ Weyl dipole can not be annihilated when meeting together. As a novel topological state, the topological Weyl-dipole phase (TWDP) has garnered significant attention, yet its realization in crystalline materials remains a challenge. Here, through first-principles calculations and theoretical analysis, we demonstrate the existence of the Weyl-dipole phase in the phonon spectra of the $P6_3$ type Y(OH)$_3$. Particularly, the Weyl dipole in this system is protected by a quantized quadrupole moment, and it distinguished from conventional Weyl dipole, as it comprises an unconventional charge-3 WP with ${\cal{C}}=-3$ and three conventional charge-1 WPs with ${\cal{C}}=1$. Consequently, the Weyl-dipole phase in Y(OH)$_3$ features unique two-dimensional (2D) sextuple-helicoid Fermi-arc states on the top and bottom surfaces, protected by the Chern number, as well as one-dimensional (1D) hinge states that connect the two Weyl dipoles along the side hinges, guaranteed by the quantized quadrupole moment. Our findings not only introduce a novel higher-order topological phase, but also promote Y(OH)$_3$ as a promising platform for exploring multi-dimensional boundaries and the interaction between first-order and second-order topologies.

cond-mat.mtrl-sci↗

arXiv2025

BiLD: Bi-directional Logits Difference Loss for Large Language Model Distillation

In recent years, large language models (LLMs) have shown exceptional capabilities across various natural language processing (NLP) tasks. However, such impressive performance often comes with the trade-off of an increased parameter size, posing significant challenges for widespread deployment. Knowledge distillation (KD) provides a solution by transferring knowledge from a large teacher model to a smaller student model. In this paper, we explore the task-specific distillation of LLMs at the logit level. Our investigation reveals that the logits of fine-tuned LLMs exhibit a more extreme long-tail distribution than those from vision models, with hidden "noise" in the long tail affecting distillation performance. Furthermore, existing logits distillation methods often struggle to effectively utilize the internal ranking information from the logits. To address these, we propose the Bi-directional Logits Difference (BiLD) loss. The BiLD loss filters out the long-tail noise by utilizing only top-$k$ teacher and student logits, and leverages the internal logits ranking information by constructing logits differences. To evaluate BiLD loss, we conduct comprehensive experiments on 13 datasets using two types of LLMs. Our results show that the BiLD loss, with only the top-8 logits, outperforms supervised fine-tuning (SFT), vanilla KL loss, and five other distillation methods from both NLP and CV fields.

cs.CL↗

arXiv2025

IGNN-Solver: A Graph Neural Solver for Implicit Graph Neural Networks

Implicit graph neural networks (IGNNs), which exhibit strong expressive power with a single layer, have recently demonstrated remarkable performance in capturing long-range dependencies (LRD) in underlying graphs while effectively mitigating the over-smoothing problem. However, IGNNs rely on computationally expensive fixed-point iterations, which lead to significant speed and scalability limitations, hindering their application to large-scale graphs. To achieve fast fixed-point solving for IGNNs, we propose a novel graph neural solver, IGNN-Solver, which leverages the generalized Anderson Acceleration method, parameterized by a tiny GNN, and learns iterative updates as a graph-dependent temporal process. To improve effectiveness on large-scale graph tasks, we further integrate sparsification and storage compression methods, specifically tailored for the IGNN-Solver, into its design. Extensive experiments demonstrate that the IGNN-Solver significantly accelerates inference on both small- and large-scale tasks, achieving a $1.5\times$ to $8\times$ speedup without sacrificing accuracy. This advantage becomes more pronounced as the graph scale grows, facilitating its large-scale deployment in real-world applications. The code to reproduce our results is available at https://github.com/landrarwolf/IGNN-Solver.

cs.LG↗

arXiv2025

Federated Neural Nonparametric Point Processes

Temporal point processes (TPPs) are effective for modeling event occurrences over time, but they struggle with sparse and uncertain events in federated systems, where privacy is a major concern. To address this, we propose \textit{FedPP}, a Federated neural nonparametric Point Process model. FedPP integrates neural embeddings into Sigmoidal Gaussian Cox Processes (SGCPs) on the client side, which is a flexible and expressive class of TPPs, allowing it to generate highly flexible intensity functions that capture client-specific event dynamics and uncertainties while efficiently summarizing historical records. For global aggregation, FedPP introduces a divergence-based mechanism that communicates the distributions of SGCPs' kernel hyperparameters between the server and clients, while keeping client-specific parameters local to ensure privacy and personalization. FedPP effectively captures event uncertainty and sparsity, and extensive experiments demonstrate its superior performance in federated settings, particularly with KL divergence and Wasserstein distance-based global aggregation.

cs.LG↗

arXiv2025

Socratic Questioning: Learn to Self-guide Multimodal Reasoning in the Wild

Complex visual reasoning remains a key challenge today. Typically, the challenge is tackled using methodologies such as Chain of Thought (COT) and visual instruction tuning. However, how to organically combine these two methodologies for greater success remains unexplored. Also, issues like hallucinations and high training cost still need to be addressed. In this work, we devise an innovative multi-round training and reasoning framework suitable for lightweight Multimodal Large Language Models (MLLMs). Our self-questioning approach heuristically guides MLLMs to focus on visual clues relevant to the target problem, reducing hallucinations and enhancing the model's ability to describe fine-grained image details. This ultimately enables the model to perform well in complex visual reasoning and question-answering tasks. We have named this framework Socratic Questioning(SQ). To facilitate future research, we create a multimodal mini-dataset named CapQA, which includes 1k images of fine-grained activities, for visual instruction tuning and evaluation, our proposed SQ method leads to a 31.2% improvement in the hallucination score. Our extensive experiments on various benchmarks demonstrate SQ's remarkable capabilities in heuristic self-questioning, zero-shot visual reasoning and hallucination mitigation. Our model and code will be publicly available.

cs.CV↗

arXiv2025

Distillation Learning Guided by Image Reconstruction for One-Shot Medical Image Segmentation

Traditional one-shot medical image segmentation (MIS) methods use registration networks to propagate labels from a reference atlas or rely on comprehensive sampling strategies to generate synthetic labeled data for training. However, these methods often struggle with registration errors and low-quality synthetic images, leading to poor performance and generalization. To overcome this, we introduce a novel one-shot MIS framework based on knowledge distillation, which allows the network to directly 'see' real images through a distillation process guided by image reconstruction. It focuses on anatomical structures in a single labeled image and a few unlabeled ones. A registration-based data augmentation network creates realistic, labeled samples, while a feature distillation module helps the student network learn segmentation from these samples, guided by the teacher network. During inference, the streamlined student network accurately segments new images. Evaluations on three public datasets (OASIS for T1 brain MRI, BCV for abdomen CT, and VerSe for vertebrae CT) show superior segmentation performance and generalization across different medical image datasets and modalities compared to leading methods. Our code is available at https://github.com/NoviceFodder/OS-MedSeg.

eess.IV↗

arXiv2025

Textual and Visual Prompt Fusion for Image Editing via Step-Wise Alignment

The use of denoising diffusion models is becoming increasingly popular in the field of image editing. However, current approaches often rely on either image-guided methods, which provide a visual reference but lack control over semantic consistency, or text-guided methods, which ensure alignment with the text guidance but compromise visual quality. To resolve this issue, we propose a framework that integrates a fusion of generated visual references and text guidance into the semantic latent space of a \textit{frozen} pre-trained diffusion model. Using only a tiny neural network, our framework provides control over diverse content and attributes, driven intuitively by the text prompt. Compared to state-of-the-art methods, the framework generates images of higher quality while providing realistic editing effects across various benchmark datasets.

cs.CV↗

arXiv2024

Task Diversity in Bayesian Federated Learning: Simultaneous Processing of Classification and Regression

This work addresses a key limitation in current federated learning approaches, which predominantly focus on homogeneous tasks, neglecting the task diversity on local devices. We propose a principled integration of multi-task learning using multi-output Gaussian processes (MOGP) at the local level and federated learning at the global level. MOGP handles correlated classification and regression tasks, offering a Bayesian non-parametric approach that naturally quantifies uncertainty. The central server aggregates the posteriors from local devices, updating a global MOGP prior redistributed for training local models until convergence. Challenges in performing posterior inference on local devices are addressed through the Pólya-Gamma augmentation technique and mean-field variational inference, enhancing computational efficiency and convergence rate. Experimental results on both synthetic and real data demonstrate superior predictive performance, OOD detection, uncertainty calibration and convergence rate, highlighting the method's potential in diverse applications. Our code is publicly available at https://github.com/JunliangLv/task_diversity_BFL.

cs.LG↗

arXiv2024

Nonstationary Sparse Spectral Permanental Process

Existing permanental processes often impose constraints on kernel types or stationarity, limiting the model's expressiveness. To overcome these limitations, we propose a novel approach utilizing the sparse spectral representation of nonstationary kernels. This technique relaxes the constraints on kernel types and stationarity, allowing for more flexible modeling while reducing computational complexity to the linear level. Additionally, we introduce a deep kernel variant by hierarchically stacking multiple spectral feature mappings, further enhancing the model's expressiveness to capture complex patterns in data. Experimental results on both synthetic and real-world datasets demonstrate the effectiveness of our approach, particularly in scenarios with pronounced data nonstationarity. Additionally, ablation studies are conducted to provide insights into the impact of various hyperparameters on model performance.

stat.ML↗