arXiv subjects

Huan Hu

Publications and source records attributed to Huan Hu.

At least 19 recordsLinked to original sources

arXiv2026

Chain-of-Trajectories: Unlocking the Intrinsic Generative Optimality of Diffusion Models via Graph-Theoretic Planning

Diffusion models operate in a reflexive System 1 mode, constrained by a fixed, content-agnostic sampling schedule. This rigidity arises from the curse of state dimensionality, where the combinatorial explosion of possible states in the high-dimensional noise manifold renders explicit trajectory planning intractable and leads to systematic computational misallocation. To address this, we introduce Chain-of-Trajectories (CoTj), a train-free framework enabling System 2 deliberative planning. Central to CoTj is Diffusion DNA, a low-dimensional signature that quantifies per-stage denoising difficulty and serves as a proxy for the high-dimensional state space, allowing us to reformulate sampling as graph planning on a directed acyclic graph. Through a Predict-Plan-Execute paradigm, CoTj dynamically allocates computational effort to the most challenging generative phases. Experiments across multiple generative models demonstrate that CoTj discovers context-aware trajectories, improving output quality and stability while reducing redundant computation. This work establishes a new foundation for resource-aware, planning-based diffusion modeling. The code is available at https://github.com/UnicomAI/CoTj.

cs.LG↗

arXiv2026

Beyond Geometry: Artistic Disparity Synthesis for Immersive 2D-to-3D

Current 2D-to-3D conversion methods achieve geometric accuracy but are artistically deficient, failing to replicate the immersive and emotionally resonant experience of professional 3D cinema. This is because geometric reconstruction paradigms mistake deliberate artistic intent, such as strategic zero-plane shifts for pop-out effects and local depth sculpting, for data noise or ambiguity. This paper argues for a new paradigm: Artistic Disparity Synthesis, shifting the goal from physically accurate disparity estimation to artistically coherent disparity synthesis. We propose Art3D, a preliminary framework exploring this paradigm. Art3D uses a dual-path architecture to decouple global depth parameters (macro-intent) from local artistic effects (visual brushstrokes) and learns from professional 3D film data via indirect supervision. We also introduce a preliminary evaluation method to quantify cinematic alignment. Experiments show our approach demonstrates potential in replicating key local out-of-screen effects and aligning with the global depth styles of cinematic 3D content, laying the groundwork for a new class of artistically-driven conversion tools.

cs.CV↗

arXiv2025

A Bio-Inspired Whisker Sensor toward Underwater Flow Sensing in Darkness and Turbidity

Underwater flow sensing is critical for unmanned underwater vehicles (UUVs) and environmental monitoring, yet existing sensors often suffer from low responsiveness, high detection thresholds, limited directional discrimination, complex packaging, and poor long-term stability, especially for navigation and target perception in turbid and cluttered waters. Previous solutions based on traditional strain gauges with limited detection accuracy or doped silicon sensors with limited detection height have shown feasibility but still face challenges in scalability, robustness under harsh aquatic conditions, and calibration complexity. This work presents a bio-inspired whisker sensor that provides a balanced solution by embedding high-gauge-factor silicon strain gauges into a flexible PDMS base, mimicking seal whiskers to offer both high sensitivity and simplified packaging. The device exhibits a linear force-resistance response with a limit of detection of 0.27 mN, maintains stability after 10,000 loading cycles, and shows minimal offset drift of less than 2 percent. It also demonstrates frequency matching in underwater dipole tests with clear longitudinal and transverse spatial response patterns. These results indicate a robust and scalable route for underwater flow sensing on UUV platforms in practical deployments.

eess.SP↗

arXiv2025

PSTF-AttControl: Per-Subject-Tuning-Free Personalized Image Generation with Controllable Face Attributes

Recent advancements in personalized image generation have significantly improved facial identity preservation, particularly in fields such as entertainment and social media. However, existing methods still struggle to achieve precise control over facial attributes in a per-subject-tuning-free (PSTF) way. Tuning-based techniques like PreciseControl have shown promise by providing fine-grained control over facial features, but they often require extensive technical expertise and additional training data, limiting their accessibility. In contrast, PSTF approaches simplify the process by enabling image generation from a single facial input, but they lack precise control over facial attributes. In this paper, we introduce a novel, PSTF method that enables both precise control over facial attributes and high-fidelity preservation of facial identity. Our approach utilizes a face recognition model to extract facial identity features, which are then mapped into the $W^+$ latent space of StyleGAN2 using the e4e encoder. We further enhance the model with a Triplet-Decoupled Cross-Attention module, which integrates facial identity, attribute features, and text embeddings into the UNet architecture, ensuring clean separation of identity and attribute information. Trained on the FFHQ dataset, our method allows for the generation of personalized images with fine-grained control over facial attributes, while without requiring additional fine-tuning or training data for individual identities. We demonstrate that our approach successfully balances personalization with precise facial attribute control, offering a more efficient and user-friendly solution for high-quality, adaptable facial image synthesis. The code is publicly available at https://github.com/UnicomAI/PSTF-AttControl.

cs.CV↗

arXiv2025

Fuzzy Reasoning Chain (FRC): An Innovative Reasoning Framework from Fuzziness to Clarity

With the rapid advancement of large language models (LLMs), natural language processing (NLP) has achieved remarkable progress. Nonetheless, significant challenges remain in handling texts with ambiguity, polysemy, or uncertainty. We introduce the Fuzzy Reasoning Chain (FRC) framework, which integrates LLM semantic priors with continuous fuzzy membership degrees, creating an explicit interaction between probability-based reasoning and fuzzy membership reasoning. This transition allows ambiguous inputs to be gradually transformed into clear and interpretable decisions while capturing conflicting or uncertain signals that traditional probability-based methods cannot. We validate FRC on sentiment analysis tasks, where both theoretical analysis and empirical results show that it ensures stable reasoning and facilitates knowledge transfer across different model scales. These findings indicate that FRC provides a general mechanism for managing subtle and ambiguous expressions with improved interpretability and robustness.

cs.CL↗

arXiv2025

Hierarchical Deep Fusion Framework for Multi-dimensional Facial Forgery Detection -- The 2024 Global Deepfake Image Detection Challenge

The proliferation of sophisticated deepfake technology poses significant challenges to digital security and authenticity. Detecting these forgeries, especially across a wide spectrum of manipulation techniques, requires robust and generalized models. This paper introduces the Hierarchical Deep Fusion Framework (HDFF), an ensemble-based deep learning architecture designed for high-performance facial forgery detection. Our framework integrates four diverse pre-trained sub-models, Swin-MLP, CoAtNet, EfficientNetV2, and DaViT, which are meticulously fine-tuned through a multi-stage process on the MultiFFDI dataset. By concatenating the feature representations from these specialized models and training a final classifier layer, HDFF effectively leverages their collective strengths. This approach achieved a final score of 0.96852 on the competition's private leaderboard, securing the 20th position out of 184 teams, demonstrating the efficacy of hierarchical fusion for complex image classification tasks.

cs.CV↗

arXiv2025

iLearnRobot: An Interactive Learning-Based Multi-Modal Robot with Continuous Improvement

It is crucial that robots' performance can be improved after deployment, as they are inherently likely to encounter novel scenarios never seen before. This paper presents an innovative solution: an interactive learning-based robot system powered by a Multi-modal Large Language Model(MLLM). A key feature of our system is its ability to learn from natural dialogues with non-expert users. We also propose chain of question to clarify the exact intent of the question before providing an answer and dual-modality retrieval modules to leverage these interaction events to avoid repeating same mistakes, ensuring a seamless user experience before model updates, which is in contrast to current mainstream MLLM-based robotic systems. Our system marks a novel approach in robotics by integrating interactive learning, paving the way for superior adaptability and performance in diverse environments. We demonstrate the effectiveness and improvement of our method through experiments, both quantitively and qualitatively.

cs.HC↗

arXiv2025

SLearnLLM: A Self-Learning Framework for Efficient Domain-Specific Adaptation of Large Language Models

When using supervised fine-tuning (SFT) to adapt large language models (LLMs) to specific domains, a significant challenge arises: should we use the entire SFT dataset for fine-tuning? Common practice often involves fine-tuning directly on the entire dataset due to limited information on the LLM's past training data. However, if the SFT dataset largely overlaps with the model's existing knowledge, the performance gains are minimal, leading to wasted computational resources. Identifying the unknown knowledge within the SFT dataset and using it to fine-tune the model could substantially improve the training efficiency. To address this challenge, we propose a self-learning framework for LLMs inspired by human learning pattern. This framework takes a fine-tuning (SFT) dataset in a specific domain as input. First, the LLMs answer the questions in the SFT dataset. The LLMs then objectively grade the responses and filter out the incorrectly answered QA pairs. Finally, we fine-tune the LLMs based on this filtered QA set. Experimental results in the fields of agriculture and medicine demonstrate that our method substantially reduces training time while achieving comparable improvements to those attained with full dataset fine-tuning. By concentrating on the unknown knowledge within the SFT dataset, our approach enhances the efficiency of fine-tuning LLMs.

cs.CL↗

arXiv2025

Ultralow-pressure mechanical-motion switching of ferroelectric polarization

Ferroelectric polarization switching, achieved by mechanical forces, enables the storage of stress information in ferroelectrics, and holds promise for human-interfacing applications. The prevailing mechanical approach is locally induced flexoelectricity with large strain gradients. However, this approach usually requires huge mechanical pressures, which greatly impedes device applications. Here, we report an approach of using triboelectric effect to mechanically, reversibly switch ferroelectric polarization across {\alpha}-In2Se3 ferroelectric memristors. Through contact electrification and electrostatic induction effects, triboelectric units are used to sensitively detect mechanical forces and generate electrical voltage pulses to trigger {\alpha}-In2Se3 resistance switching. We realize multilevel resistance states under different mechanical forces, by which a neuromorphic stress system is demonstrated. Strikingly, we achieve the reversal of {\alpha}-In2Se3 ferroelectric polarization with a record-low mechanical pressure of ~ 10 kPa, and even with tactile touches. Our work provides a fundamental but pragmatic strategy for creating mechanical-tactile ferroelectric memory devices.

cond-mat.mtrl-sci↗

arXiv2025

A Multimodal Benchmark Dataset and Model for Crop Disease Diagnosis

While conversational generative AI has shown considerable potential in enhancing decision-making for agricultural professionals, its exploration has predominantly been anchored in text-based interactions. The evolution of multimodal conversational AI, leveraging vast amounts of image-text data from diverse sources, marks a significant stride forward. However, the application of such advanced vision-language models in the agricultural domain, particularly for crop disease diagnosis, remains underexplored. In this work, we present the crop disease domain multimodal (CDDM) dataset, a pioneering resource designed to advance the field of agricultural research through the application of multimodal learning techniques. The dataset comprises 137,000 images of various crop diseases, accompanied by 1 million question-answer pairs that span a broad spectrum of agricultural knowledge, from disease identification to management practices. By integrating visual and textual data, CDDM facilitates the development of sophisticated question-answering systems capable of providing precise, useful advice to farmers and agricultural professionals. We demonstrate the utility of the dataset by finetuning state-of-the-art multimodal models, showcasing significant improvements in crop disease diagnosis. Specifically, we employed a novel finetuning strategy that utilizes low-rank adaptation (LoRA) to finetune the visual encoder, adapter and language model simultaneously. Our contributions include not only the dataset but also a finetuning strategy and a benchmark to stimulate further research in agricultural technology, aiming to bridge the gap between advanced AI techniques and practical agricultural applications. The dataset is available at https: //github.com/UnicomAI/UnicomBenchmark/tree/main/CDDMBench.

cs.CV↗

arXiv2025

Optimizing for the Shortest Path in Denoising Diffusion Model

In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF.

cs.CV↗

arXiv2025

Correction to the quantum relation of photons in the Doppler effect based on a special Lorentz violation model

The possibility of the breaking of Lorentz symmetry has been discussed in many models of quantum gravity. In this paper we follow the Lorentz violation model in Ref. [1] (i.e., our previous work) to discuss the Doppler frequency shift of photons and the Compton scattering process between photons and electrons, pointing out that following the idea in Ref. [1] we have to modify the usual quantum relation of photons in the Doppler effect. But due to the current limited information and knowledge, we could not yet determine the specific expression for the correction coefficient in the modified quantum relation of photons. However, the phenomenon called spontaneous radiation in a cyclotron maser give us an opportunity to see what the expression for this correction coefficient might look like. Therefor, under some necessary constraints, we construct a very concise expression for this correction coefficient through the discussion of different cases. And then we use this expression to analyze the wavelength of radiation in the cyclotron maser, which tends to a limited value at v is close to c, rather than to 0 as predicted by the Lorentz model. And the inverse Compton scattering phenomenon is also discussed and we find that there is a limit to the maximum energy that can be obtained by photons in the collision between extremely relativistic particles and low-energy photons, which conclusion is also very different from that obtained from the Lorentz model, in which the energy that can be obtained by the photon tends to be infinite as the velocity of particle is close to c. This paper still follows the purpose in Ref. [1] that the energy and momentum of particles (i.e., any particles, including photons) cannot be infinite, otherwise it will make some physical scenarios invalid.

physics.gen-ph↗

arXiv2024

Training Interactive Agent in Large FPS Game Map with Rule-enhanced Reinforcement Learning

In the realm of competitive gaming, 3D first-person shooter (FPS) games have gained immense popularity, prompting the development of game AI systems to enhance gameplay. However, deploying game AI in practical scenarios still poses challenges, particularly in large-scale and complex FPS games. In this paper, we focus on the practical deployment of game AI in the online multiplayer competitive 3D FPS game called Arena Breakout, developed by Tencent Games. We propose a novel gaming AI system named Private Military Company Agent (PMCA), which is interactable within a large game map and engages in combat with players while utilizing tactical advantages provided by the surrounding terrain. To address the challenges of navigation and combat in modern 3D FPS games, we introduce a method that combines navigation mesh (Navmesh) and shooting-rule with deep reinforcement learning (NSRL). The integration of Navmesh enhances the agent's global navigation capabilities while shooting behavior is controlled using rule-based methods to ensure controllability. NSRL employs a DRL model to predict when to enable the navigation mesh, resulting in a diverse range of behaviors for the game AI. Customized rewards for human-like behaviors are also employed to align PMCA's behavior with that of human players.

cs.AI↗

arXiv2024

Viscous universe with cosmological constant

We investigated a bulk viscous fluid universe with cosmological constant {\Lambda} by assuming that the bulk viscosity to be proportional to the Hubble parameter. We found that for an expanding universe, the (relative) matter density will be always greater than a non-zero constant, and tends to this non-zero constant in the future. We show that the bulk viscosity model has a significantly better fitting to the combined SNeIa + CMB + BAO + H(z) data than the {\Lambda}CDM model. Generally, the evolution or values of some cosmological parameters predicted by the bulk viscosity model do not deviate significantly from which are obtained from the {\Lambda}CDM model since the bulk viscosity coefficient obtained from the astronomical observational data is so small. We also made a statefinder analysis of the bulk viscosity model and found that the evolution of the {r, s} parameters behaves in such a way that 0 < s < 1, 0.945 < r <1, indicating the bulk viscosity model is different from the {\Lambda}CDM model.

gr-qc↗

arXiv2023

A Special Lorentz Violation Model and a Special Rainbow Function

In order to characterize the common feature of the general Lorentz violation models that the local speed of light is variable at ultrahigh energy scale, we introduced a parameter n to characterize the variation of the speed of light between inertial systems. And in order not to violate some fundamental principles and experiments' results, some constraints were imposed on n. Then a coordinate transformation with the parameter n, which meets the symmetry of inertial systems, was naturally obtained. But just to satisfy the symmetry of inertial systems, the expression for n cannot be determined. Inspired by the idea of DSR model, we then discussed a specific expression for n that makes the particle's energy have a limit rather than be infinite derived from the Lorentz model. We found that the modified dispersion relation corresponding to the specific expression for n is deeply associated with the general formula of dispersion relation from the DSR. The motivation of introducing such a parameter n was also discussed. Finally, we discussed the possible LIV effects from astrophysical observations in GRB events and photon annihilation events. Using the data from GRB 190114C we investigated the special dispersion relation obtained in this paper and found that it appears to fit these data better than the two models corresponding to the first or second order approximation of the general formula of dispersion relation. And similar to the other Lorentz violation models corresponding to the subluminal case, the modified dispersion relation obtained in this paper can also result in increasing the energy threshold of photon annihilation reaction, which is expected to be verified by the future energetic photons events.

physics.gen-ph↗

arXiv2023

Enhanced photovoltaic effect in graphene-silicon Schottky junction under mechanical manipulation

Graphene-silicon Schottky junction (GSJ) which has the potential for large-scale manufacturing and integration can bring new opportunities to Schottky solar cells for photovoltaic (PV) power conversion. However, the essential power conversion limitation for these devices lies in the small open-circuit voltage ($V_{oc}$), which depends on the Schottky barrier height (SBH). In this study, we introduce an electromechanical method based on the flexoelectric effect to enhance the PV efficiency in GSJ. By atomic force microscope (AFM) tip-based indentation and in situ current measurement, the current-voltage (I-V) responses under flexoelectric strain gradient are obtained. The $V_{oc}$ is observed to increase for up to 20$\%$, leading to an evident improvement of the power conversion efficiency. Our studies suggest that strain gradient may offer unprecedented opportunities for the development of GSJ based flexo-photovoltaic applications.

cond-mat.mtrl-sci↗

arXiv2022

Charge Crowding in Graphene-Silicon Diodes

The performance of nanoscale electronic devices based on a two-three dimensional (2D-3D) interface is significantly affected by the electrical contacts that interconnect these materials with external circuitry. This work investigates charge transport effects at the 2D-3D ohmic contact coupled with the thermionic injection model for graphene/Si Schottky junction. Here, w e focus on the intrinsic properties of graphene-metal contacts, paying particular attention to the nature of the contact failure mechanism under high electrical stress. According to our findings, severe current crowding (CC) effects in highly conductive electrical contact significantly affect device failure that can be reduced by spatially varying the contact properties and geometry. The impact of electrical breakdown on material degradation is systematically analyzed by atomic force, Raman, scanning electron, and energy dispersive X-ray spectroscopies. Our devices withstand high electrostatic discharge spikes over a longer period, manifesting high robustness and operational stability. This research paves the way towards a highly robust and reliable graphene/Si heterostructure in futuristic on-chip integration in dynamic switching. The methods we employed here can be extended for other nanoscale electronic devices based on 2D-3D interfaces

physics.app-ph↗

arXiv2022

Data-Centric AI Paradigm Based on Application-Driven Fine-Grained Dataset Design

Deep learning has a wide range of applications in industrial scenario, but reducing false alarm (FA) remains a major difficulty. Optimizing network architecture or network parameters is used to tackle this challenge in academic circles, while ignoring the essential characteristics of data in application scenarios, which often results in increased FA in new scenarios. In this paper, we propose a novel paradigm for fine-grained design of datasets, driven by industrial applications. We flexibly select positive and negative sample sets according to the essential features of the data and application requirements, and add the remaining samples to the training set as uncertainty classes. We collect more than 10,000 mask-wearing recognition samples covering various application scenarios as our experimental data. Compared with the traditional data design methods, our method achieves better results and effectively reduces FA. We make all contributions available to the research community for broader use. The contributions will be available at https://github.com/huh30/OpenDatasets.

cs.CV↗