Author name cluster

Zhenyu Wen

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

5 papers

2 author rows

AAAI Conference 2026 Conference Paper

Beyond Single-Point Perturbation: A Hierarchical, Manifold-Aware Approach to Diffusion Attacks

Zhijie Wang
Lin Wang
Zhenyu Wen
Cong Wang

Latent Diffusion Models have become a powerful tool for generating high-fidelity unrestricted adversarial examples. However, the existing methods typically perturb only the initial latent or rely on prompt engineering, which is ill-suited to the iterative nature of the diffusion process, plus optimization instability due to external text prompts and cumulative drift that push the adversarial images off the data manifold. In this paper, we propose a hierarchical attack framework that operates in alignment with the model's generative manifold and leverages intermediate denoising states to maximize attack transferability and visual fidelity. Extensive experiments show that the proposed attack improves adversarial transferability by 10-20% against a diverse set of normally-trained models and achieves over 10.5% higher success rate against adversarially-defended models, while simultaneously enhancing visual quality by 1.0-1.2 FID reduction and 16.7% LPIPS improvements.

PDF Details DOI

AAAI Conference 2026 Conference Paper

FIRM-MoE:Fine-GrainedExpert Decomposition for Resource-Adaptive MoE Inference

Keyu Chen
Qihang Zhou
Bin Qian
Zhenyu Wen
Wenchao Meng
Shibo He

Mixture-of-Experts (MoE) is a sparse neural architecture that significantly increases model capacity while maintaining low computational complexity. However, deploying MoE-based large language models (LLMs) on memory-constrained edge devices remains challenging due to their substantial memory requirements. To address this issue, we propose FIRM-MoE, a fine-grained expert offloading framework designed to enable flexible and efficient MoE inference. The core insight of our approach is to reduce the risk of inaccurate expert loading by decomposing each expert into fine-grained sub-experts and then dynamically allocating them through a fine-grained scheduling strategy. To further reduce the error in expert loading, we introduce a multi-layer expert prediction mechanism and a resource-adaptive expert pre-loading algorithm to enable more robust expert allocation. This design allows our model to achieve more efficient expert utilization and improved resilience to prediction errors. We conduct extensive experiments to demonstrate the superiority of FIRM-MoE across diverse memory constraints. The results show that FIRM-MoE achieves up to 1.5× speedup and 2.8× memory savings in decoding, compared to state-of-the-art MoE offloading strategies.

PDF Details DOI

AAAI Conference 2026 Conference Paper

RFF-TTA: Physical Information-Aware Prototype for Temporally Varying RF Fingerprinting Online Test-Time-Adaptation

Taotao Li
YiYang Li
Zhenyu Wen
Jiahao Lin
Jinhao Wan
Jie Su
Cong Wang
Zhen Hong

In recent years, RF fingerprinting (RFF) has emerged as a promising technology for wireless device authentication. However, temporal variations in device load and temperature, along with channel effects, lead to inconsistencies in RFF distributions between training and testing phases. As a result, deep learning (DL)-based recognition models often suffer from degraded performance. To address this problem, we propose the first test-time-adaptation (TTA) approach to improve the domain generalization ability of RFF recognition models. We first analyze the causes of time-varying RFF distribution shifts, such as carrier frequency offset (CFO), and develop a physical impairment-based data augmentation strategy. Based on this, we further propose a physically information-aware prototype to guide the model for TTA. Our method requires no model retraining or labeled test samples, and is a lightweight, nonparametric solution. Finally, our approach is extensively evaluated using mobile phones with the IEEE 802.11 orthogonal frequency division multiplexing (OFDM) system, which demonstrates that our scheme can effectively improve RFF average recognition performance by about 7.8%.

PDF Details DOI

IS Journal 2024 Journal Article

Unraveling Complexity: An Exploration Into Large-Scale Multimodal Signal Processing

Zhenyu Wen
Yuheng Ye
Jie Su
Taotao Li
Jinhao Wan
Shilian Zheng
Zhen Hong
Shibo He

Advanced communication systems and military reconnaissance are increasingly prevalent in high-tech environments, greatly supported by the flourishing in signal processing technologies. The recent exponential proliferation of sensors led to an unprecedented expansion in the scale and diversity of signals across various modalities. Such an influx poses significant challenges in effectively integrating multimodal signal data to deliver comprehensive and interpretive solutions across a diverse range of applications. In this article, we provide an overview of the core issues, challenges, and future research directions in different stages of developing large-scale multimodal signal processing models. Additionally, we introduce a prior investigation into signal representation learning, where we propose a contrastive-learning-based framework to extract fine-grained signal features under few-shot conditions. Our proposed framework achieves a 24. 1% performance improvement over baseline approaches, consistently demonstrating superiority over state-of-the-art methods. The code is accessible in this repository: https://github.com/YYH211/LSM.

Details DOI

ICRA Conference 2023 Conference Paper

Toward Cooperative 3D Object Reconstruction with Multi-agent

Xiong Li
Zhenyu Wen
Leiqiang Zhou
Chenwei Li
Yejian Zhou
Taotao Li
Zhen Hong

We study the problem of object reconstruction in a multi-agent collaboration scenario. Specifically, we focus on the reconstruction of specific goals through several cooperative agents equipped with vision sensors to achieve higher efficiency than single agents. Our main insight is that a complete 3D object can be split into several local 3D models and assigned to different agents. In addition, we can use the salient characteristics of the collaboration agent itself to help realize the integration of local models. We develop a novel pipeline that first restores local 3D models from the images obtained from different agents, then the relative poses between collaborative agents are estimated by aligning intrinsic features. After that, all local models are integrated using the estimated parameters. Extensive experiments show that our proposed method is capable of accurately reconstructing 3D objects in the real world in a multi-agent collaborative manner. The full reconstruction pipeline is released to the public as an open-source project.

Details