arXiv subjects

Zhengdong Hu

Publications and source records attributed to Zhengdong Hu.

13 recordsLinked to original sources

arXiv2026

Replication in Visual Diffusion Models: A Survey and Outlook

Visual diffusion models have revolutionized the field of creative AI, producing high-quality and diverse content. However, they inevitably memorize training images or videos, subsequently replicating their concepts, content, or styles during inference. This phenomenon raises significant concerns about privacy, security, and copyright within generated outputs. In this survey, we provide the first comprehensive review of replication in visual diffusion models, marking a novel contribution to the field by systematically categorizing the existing studies into unveiling, understanding, and mitigating this phenomenon. Specifically, unveiling mainly refers to the methods used to detect replication instances. Understanding involves analyzing the underlying mechanisms and factors that contribute to this phenomenon. Mitigation focuses on developing strategies to reduce or eliminate replication. Beyond these aspects, we also review papers focusing on its real-world influence. For instance, in the context of healthcare, replication is critically worrying due to privacy concerns related to patient data. Finally, the paper concludes with a discussion of the ongoing challenges, such as the difficulty in detecting and benchmarking replication, and outlines future directions including the development of more robust mitigation techniques. By synthesizing insights from diverse studies, this paper aims to equip researchers and practitioners with a deeper understanding at the intersection between AI technology and social good. We release this project at https://github.com/WangWenhao0716/Awesome-Diffusion-Replication.

cs.CV↗

arXiv2026

PixelEyes: Decoupling Perception and Reasoning for Pinpoint Visual Evidence Seeking

This paper explores multi-turn visual reasoning and observes that MLLMs repeatedly fail to localize the target, leading to long, redundant trajectories. We attribute this failure to the entanglement of reasoning and perception within a single model, the MLLM reasons and localizes simultaneously, and inaccurate localization triggers additional reasoning turns that bloat the trajectory. To solve this problem, we propose PixelEyes, a multi-turn visual reasoning agent that explicitly decouples reasoning from perception, i.e., the reasoner decides what to look for, while a specialized perception tool answers where it is. Specifically, PixelEyes introduces 1) Mask-guided Visual Search. A referring segmentation model is invoked to provide mask-precise localization, freeing the reasoner from the need to compensate for imprecise grounding. 2) Semantic-region Breadth-first Search (BFS). To eliminate redundant loops caused by repeatedly cropping incorrect sub-regions, we organize exploration as a breadth-first search over semantic regions. To internalize these capabilities, we construct the PixelEyes-6K dataset by resynthesizing expert trajectories from existing data. This explicitly embeds our mask-guided search and BFS logic into the model. We further introduce Pinpoint-Bench, a zero-hint visual search benchmark, i.e., no location cues are provided in the question, with instance-level masks and bounding boxes that separate localization failures from reasoning failures, enabling fine-grained analysis of failure modes such as inattentional blindness. Recent state-of-the-art MLLMs and visual reasoning agents leave large headroom on Pinpoint-Bench, demonstrating its quality and difficulty. Code and models are open-sourced.

cs.CV↗

arXiv2026

Skill-3D: Evolving Scene-Aware Skills for Agentic 3D Spatial Reasoning

This paper explores agentic 3D spatial understanding, i.e., MLLM agents performing 3D reasoning through tool use. Existing methods often misuse tools and exhibit biased tool preferences under 3D scenarios, leaving the agentic paradigm with only marginal gains over non-agentic strategies. We reveal that 3D spatial reasoning tasks are heterogeneous across scenes, while these agents apply a uniform tool-use strategy to all scenes rather than selecting tools according to the specific scene and task. To address this, we propose Skill-3D, a framework that learns self-evolving scene-aware skills. Specifically, Skill-3D identifies the task scene and records the agent's tool-use trajectory into a Scene Memory, where successful trajectories from similar scenes are aggregated and distilled into a reusable scene-aware skill, with failed ones attached to the skill as lessons. During training, once a similar scene recurs, the corresponding skill is injected to guide the agent, producing new trajectories whose successes and failures further refine the skill, forming a loop in which the memory and the skill library co-evolve. Experiments show that Skill-3D substantially improves tool utilization in 3D spatial reasoning (from 39% to 78% on VSI-Bench), driving the agent toward correct and sufficient tool use. For instance, it improves Gemini-3-Flash by 67% on MMSI-Bench. Furthermore, we conduct agentic post-training over skill-guided trajectories, which boosts Qwen3-VL-8B by 60% on VSI-Bench.

cs.CV↗

arXiv2026

CDiT: Conditional Diffusion Transformer for Geometry-Aware Terahertz Cross Far- and Near-Field Channel Generation

Accurate channel modeling is fundamental to design and evaluation of Terahertz (THz) ultra-massive multiple-input multiple-output (UM-MIMO) systems. However, existing model-based approaches typically rely on simplified assumptions, such as sparsity or predefined parametric structures, which are insufficient to capture the complex spatial variations and cross far-/near-field propagation characteristics of practical THz channels. In this paper, a conditional diffusion transformer (CDiT) framework is proposed for high-fidelity THz channel generation. By leveraging the state-of-the-art hybrid planar-spherical wave model (HPSM), THz channel modeling is formulated as a geometry-aware conditional generative learning problem in the sparse beamspace domain. Position information is incorporated as a conditioning signal within a diffusion-transformer architecture, enabling effective learning of the spatially dependent channel distribution. By combining the strong distribution modeling capability of diffusion models with the global dependency modeling strength of transformers, the proposed framework achieves controllable and high-fidelity THz channel synthesis. Extensive experiments on realistic THz channel datasets demonstrate that the proposed framework converges stably and significantly outperforms representative benchmark methods. The proposed framework provides a promising data-driven paradigm for THz channel modeling in next-generation wireless systems.

eess.SP↗

arXiv2026

When AI Meets Terahertz: A Survey on the Symbiosis of Artificial Intelligence and Terahertz Networks

The Terahertz (THz) band (0.1-10 THz) has emerged as a critical frontier for future communication systems, offering ultra-wide bandwidths that enable Terabits-per-second (Tbps) wireless links and high-precision sensing and imaging. However, practical deployment of THz systems is hindered by unique challenges, including intricate channel characteristics, high-dimensional and large-scale optimization problems, and highly dynamic network environments. Artificial Intelligence (AI) serves as a transformative enabler to address these challenges, providing robust capabilities for precise modeling, advanced signal processing, complex optimization, real-time decision-making, and prediction, among others. Reciprocally, the unprecedented bandwidth and high-resolution sensing capabilities of THz networks provide a promising physical infrastructure for AI, facilitating training, inference, and data collection. This survey presents a systematic and comprehensive overview of AI-driven solutions across the entire THz communication network and the symbiosis of AI and THz networks. To begin with, a foundational overview of AI technologies tailored for wireless communications is presented. Subsequently, AI-based innovations are investigated, spanning from hardware design, channel modeling, physical layer optimization, up to higher-layer network protocols and advanced THz services, including mobile edge computing and sensing-empowered applications. In parallel, the capacity of THz networks to serve AI is examined, underscoring a profound paradigm shift towards a mutual symbiosis where AI and THz co-evolve and empower each other. Finally, by synthesizing these state-of-the-art advancements and identifying open research directions, this survey highlights the potential of AI in copilot with development of THz communication systems.

eess.SP↗

arXiv2026

Choosing How to Remember: Adaptive Memory Structures for LLM Agents

Memory is critical for enabling large language model (LLM) based agents to maintain coherent behavior over long-horizon interactions. However, existing agent memory systems suffer from two key gaps: they rely on a one-size-fits-all memory structure and do not model memory structure selection as a context-adaptive decision, limiting their ability to handle heterogeneous interaction patterns and resulting in suboptimal performance. We propose a unified framework, FluxMem, that enables adaptive memory organization for LLM agents. Our framework equips agents with multiple complementary memory structures. It explicitly learns to select among these structures based on interaction-level features, using offline supervision derived from downstream response quality and memory utilization. To support robust long-horizon memory evolution, we further introduce a three-level memory hierarchy and a Beta Mixture Model-based probabilistic gate for distribution-aware memory fusion, replacing brittle similarity thresholds. Experiments on two long-horizon benchmarks, PERSONAMEM and LoCoMo, demonstrate that our method achieves average improvements of 9.18% and 6.14%.

cs.AI↗

arXiv2025

DiffPace: Diffusion-based Plug-and-play Augmented Channel Estimation in mmWave and Terahertz Ultra-Massive MIMO Systems

Millimeter-wave (mmWave) and Terahertz (THz)-band communications hold great promise in meeting the growing data-rate demands of next-generation wireless networks, offering abundant bandwidth. To mitigate the severe path loss inherent to these high frequencies and reduce hardware costs, ultra-massive multiple-input multiple-output (UM-MIMO) systems with hybrid beamforming architectures can deliver substantial beamforming gains and enhanced spectral efficiency. However, accurate channel estimation (CE) in mmWave and THz UM-MIMO systems is challenging due to high channel dimensionality and compressed observations from a limited number of RF chains, while the hybrid near- and far-field radiation patterns, arising from large array apertures and high carrier frequencies, further complicate CE. Conventional compressive sensing based frameworks rely on predefined sparsifying matrices, which cannot faithfully capture the hybrid near-field and far-field channel structures, leading to degraded estimation performance. This paper introduces DiffPace, a diffusion-based plug-and-play method for channel estimation. DiffPace uses a diffusion model (DM) to capture the channel distribution based on the hybrid spherical and planar-wave (HPSM) model. By applying the plug-and-play approach, it leverages the DM as prior knowledge, improving CE accuracy. Moreover, DM performs inference by solving an ordinary differential equation, minimizing the number of required inference steps compared with stochastic sampling method. Experimental results show that DiffPace achieves competitive CE performance, attaining -15 dB normalized mean square error (NMSE) at a signal-to-noise ratio (SNR) of 10 dB, with 90\% fewer inference steps compared to state-of-the-art schemes, simultaneously providing high estimation precision and enhanced computational efficiency.

eess.SP↗

arXiv2025

Origin Identification for Text-Guided Image-to-Image Diffusion Models

Text-guided image-to-image diffusion models excel in translating images based on textual prompts, allowing for precise and creative visual modifications. However, such a powerful technique can be misused for spreading misinformation, infringing on copyrights, and evading content tracing. This motivates us to introduce the task of origin IDentification for text-guided Image-to-image Diffusion models (ID$^2$), aiming to retrieve the original image of a given translated query. A straightforward solution to ID$^2$ involves training a specialized deep embedding model to extract and compare features from both query and reference images. However, due to visual discrepancy across generations produced by different diffusion models, this similarity-based approach fails when training on images from one model and testing on those from another, limiting its effectiveness in real-world applications. To solve this challenge of the proposed ID$^2$ task, we contribute the first dataset and a theoretically guaranteed method, both emphasizing generalizability. The curated dataset, OriPID, contains abundant Origins and guided Prompts, which can be used to train and test potential IDentification models across various diffusion models. In the method section, we first prove the existence of a linear transformation that minimizes the distance between the pre-trained Variational Autoencoder (VAE) embeddings of generated samples and their origins. Subsequently, it is demonstrated that such a simple linear transformation can be generalized across different diffusion models. Experimental results show that the proposed method achieves satisfying generalization performance, significantly surpassing similarity-based methods ($+31.6\%$ mAP), even those with generalization designs. The project is available at https://id2icml.github.io.

cs.CV↗

arXiv2024

DecTest: A Decentralised Testing Architecture for Improving Data Accuracy of Blockchain Oracle

Blockchain technology ensures secure and trustworthy data flow between multiple participants on the chain, but interoperability of on-chain and off-chain data has always been a difficult problem that needs to be solved. To solve the problem that blockchain systems cannot access off-chain data, oracle is introduced. However, existing research mainly focuses on the consistency and integrity of data, but ignores the problem that oracle nodes may be externally attacked or provide false data for selfish motives, resulting in the unresolved problem of data accuracy. In this paper, we introduce a new Decentralized Testing architecture (DecTest) that aims to improve data accuracy. A blockchain oracle random secret testing mechanism is first proposed to enhance the monitoring and verification of nodes by introducing a dynamic anonymized question-verification committee. Based on this, a comprehensive evaluation incentive mechanism is designed to incentivize honest work performance by evaluating nodes based on their reputation scores. The simulation results show that we successfully reduced the discrete entropy value of the acquired data and the real value of the data by 61.4%.

cs.CR↗

arXiv2024

Transfer Learning Enabled Transformer based Generative Adversarial Networks (TT-GAN) for Terahertz Channel Modeling and Generating

Terahertz (THz) communications, ranging from 100 GHz to 10 THz, are envisioned as a promising technology for 6G and beyond wireless systems. As foundation of designing THz communications, channel modeling and characterization are crucial to scrutinize the potential of the new spectrum. However, current channel modeling and standardization heavily rely on measurements, which are both time-consuming and costly to obtain in the THz band. Here, we propose a Transfer learning enabled Transformer based Generative Adversarial Network (TT-GAN) for THz channel modeling. Specifically, as a fundamental building block, a GAN is exploited to generate channel parameters, which can substitute measurements. To greatly improve the accuracy, the first T, i.e., a transformer structure with a self-attention mechanism is incorporated in GAN. Still incurring errors compared with ground-truth measurement, the second T, i.e., a transfer learning is designed to solve the mismatch between the formulated network and measurement. The proposed TT-GAN can achieve high accuracy in channel modeling, while requiring only rather limited amount of measurement, which is a promising complementary of channel standardization that fundamentally differs from the current techniques that heavily rely on measurement.

eess.SP↗

arXiv2023

Transformer-based GAN for Terahertz Spatial-Temporal Channel Modeling and Generating

Terahertz (THz) communications are envisioned as a promising technology for 6G and beyond wireless systems, providing ultra-broad continuous bandwidth and thus Terabit-per-second (Tbps) data rates. However, as foundation of designing THz communications, channel modeling and characterization are fundamental to scrutinize the potential of the new spectrum. Relied on time-consuming and costly physical measurements, traditional statistical channel modeling methods suffer from the problem of low accuracy with the assumed certain distributions and empirical parameters. In this paper, a transformer-based generative adversarial network modeling method (T-GAN) is proposed in the THz band, which exploits the advantage of GAN in modeling the complex distribution, and the powerful expressive capability of transformer structure. Experimental results reveal that the distribution of channels generated by the proposed T-GAN method shows good agreement with the original channels in terms of the delay spread and angular spread. Moreover, T-GAN achieves good performance in modeling the power delay angular profile, with 2.18 dB root-mean-square error (RMSE).

eess.SP↗

arXiv2023

Transfer Generative Adversarial Networks (T-GAN)-based Terahertz Channel Modeling

Terahertz (THz) communications are envisioned as a promising technology for 6G and beyond wireless systems, providing ultra-broad bandwidth and thus Terabit-per-second (Tbps) data rates. However, as foundation of designing THz communications, channel modeling and characterization are fundamental to scrutinize the potential of the new spectrum. Relied on physical measurements, traditional statistical channel modeling methods suffer from the problem of low accuracy with the assumed certain distributions and empirical parameters. Moreover, it is time-consuming and expensive to acquire extensive channel measurement in the THz band. In this paper, a transfer generative adversarial network (T-GAN) based modeling method is proposed in the THz band, which exploits the advantage of GAN in modeling the complex distribution, and the benefit of transfer learning in transferring the knowledge from a source task to improve generalization about the target task with limited training data. Specifically, to start with, the proposed GAN is pre-trained using the simulated dataset, generated by the standard channel model from 3rd generation partnerships project (3GPP). Furthermore, by transferring the knowledge and fine-tuning the pre-trained GAN, the T-GAN is developed by using the THz measured dataset with a small amount. Experimental results reveal that the distribution of PDPs generated by the proposed T-GAN method shows good agreement with measurement. Moreover, T-GAN achieves good performance in channel modeling, with 9 dB improved root-mean-square error (RMSE) and higher Structure Similarity Index Measure (SSIM), compared with traditional 3GPP method.

eess.SP↗

arXiv2022

PRINCE: A Pruned AMP Integrated Deep CNN Method for Efficient Channel Estimation of Millimeter-wave and Terahertz Ultra-Massive MIMO Systems

Millimeter-wave (mmWave) and Terahertz (THz)-band communications exploit the abundant bandwidth to fulfill the increasing data rate demands of 6G wireless communications. To compensate for the high propagation loss with reduced hardware costs, ultra-massive multiple-input multiple-output (UM-MIMO) with a hybrid beamforming structure is a promising technology in the mmWave and THz bands. However, channel estimation (CE) is challenging for hybrid UM-MIMO systems, which requires recovering the high-dimensional channels from severely few channel observations. In this paper, a Pruned Approximate Message Passing (AMP) Integrated Deep Convolutional-neural-network (DCNN) CE (PRINCE) method is firstly proposed, which enhances the estimation accuracy of the AMP method by appending a DCNN network. Moreover, by truncating the insignificant feature maps in the convolutional layers of the DCNN network, a pruning method including training with regularization, pruning and refining procedures is developed to reduce the network scale. Simulation results show that the PRINCE achieves a good trade-off between the CE accuracy and significantly low complexity, with normalized-mean-square-error (NMSE) of $-10$ dB at signal-to-noise-ratio (SNR) as $10$ dB after eliminating $80\%$ feature maps.

eess.SP↗