ZWZhengqi Wen

Zhengqi Wen

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Zhengqi Wen has a total of 31 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Label Self-Supervised Ship Detection Training Method on One-Bit Non-Reconstructed SAR Images

    by Can Su, Wei Yang, Hongcheng Zeng, Xinwei An, Danni Jiang, Zhengqi Wen, Yikui Zhai, Xiaolin Tian, Jie Chen

    One-bit quantization technology effectively simplifies the hardware architecture of synthetic aperture radar (SAR) and significantly reduces the volume of echo data, making it a pivotal research direction in spaceborne SAR systems. However, the high computational cost and latency associated with one-bit SAR signal reconstruction have limited the efficiency of downstream target detection applications. To address the demand for efficient ship target detection in one-bit SAR imagery, a label self-supervised ship detection training method on one-bit non-reconstructed SAR images is proposed in this paper, which enables generic deep learning models to perform detection directly in one-bit SAR images based on the chirp scaling algorithm (CSA) without signal recovery during inference. First, a training framework based on curriculum learning is designed, aiming to discriminate between real ship targets and false targets. Second, a one-bit reconstruction and training label self-supervised correction branch is incorporated into the training framework to autonomously generate false target labels and progressively guide the model to learn false target features. Third, a one-bit non-reconstructed image dataset is proposed to validate the trained model, which maps only reliable ship target annotations from full-precision echo imaging results. Finally, experiments conducted on real SAR data demonstrate that the detection model derived from the proposed training method can directly detect ship targets in the non-reconstructed images. Compared with the classic “reconstruction-and-detection” workflow, the trained model runs inference, bypassing the reconstruction process, and achieves a 20-fold improvement in efficiency without compromising detection accuracy. This method provides a practical solution for the rapid acquisition of target information in large-scale ship monitoring using one-bit SAR.

  • Open Access

    Decoupled visual processing: Efficient multimodal adaptation via modality-specific transformer substitution

    by Mingkuan Feng, Zhengqi Wen, Jianhua Tao

    Multimodal large language models (MLLMs) have demonstrated remarkable capabilities by integrating visual and textual understanding within a unified transformer architecture. However, fine-tuning all parameters of these models for visual instruction tuning is computationally expensive and often unnecessary, as the representation requirements for visual and textual tokens diverge significantly in the deeper layers of the network. In this paper, we propose decoupled visual processing (DVP), an efficient training framework that replaces the upper decoder layers of a pretrained LLM with a lightweight, independently trainable single transformer block dedicated exclusively to visual token processing. Specifically, after shared processing through the first half of the decoder layers, visual and textual tokens are split: visual tokens are routed through a newly initialized single transformer block while textual tokens continue through the original frozen decoder layers. The two streams are then concatenated before the language modeling head. During training, only the single transformer block is updated, dramatically reducing the number of trainable parameters. Experiments on the LLaVA-1.5 framework demonstrate that DVP achieves competitive performance on MME, POPE, and ChartQA benchmarks while training only a fraction of the total parameters, suggesting that visual representations in MLLMs can be effectively learned through a decoupled, parameter-efficient pathway.

  • Open Access

    Edit Content, Preserve Acoustics: Imperceptible Text-Based Speech Editing via Self-Consistency Rewards

    by Yong Ren, Jiangyan Yi, Jianhua Tao, Zhengqi Wen, Tao Wang

    Imperceptible text-based speech editing allows users to modify spoken content by altering the transcript. It demands that modified segments fuse seamlessly with the surrounding context. Prevalent methods operating in the acoustic space suffer from inherent content-style entanglement, leading to generation instability and boundary artifacts. In this paper, we propose a novel framework grounded in the principle of "Edit Content, Preserve Acoustics". Our approach relies on two core components: (1) Structural Foundations, which decouples editing into a stable semantic space while delegating acoustic reconstruction to a Flow Matching decoder; and (2) Perceptual Alignment, which employs a novel Self-Consistency Rewards Group Relative Policy Optimization. By leveraging a pre-trained Text-to-Speech model as an implicit critic -- complemented by strict intelligibility and duration constraints -- we effectively align the edited semantic token sequence with the original context. Empirical evaluations demonstrate that our method significantly outperforms state-of-the-art autoregressive and non-autoregressive baselines, achieving superior intelligibility, robustness, and perceptual quality.

  • Open Access

    DashFusion: Dual-stream Alignment with Hierarchical Bottleneck Fusion for Multimodal Sentiment Analysis

    by Yuhua Wen, Qifei Li, Yingying Zhou, Yingming Gao, Zhengqi Wen, Jianhua Tao, Ya Li

    Multimodal sentiment analysis (MSA) integrates various modalities, such as text, image, and audio, to provide a more comprehensive understanding of sentiment. However, effective MSA is challenged by alignment and fusion issues. Alignment requires synchronizing both temporal and semantic information across modalities, while fusion involves integrating these aligned features into a unified representation. Existing methods often address alignment or fusion in isolation, leading to limitations in performance and efficiency. To tackle these issues, we propose a novel framework called Dual-stream Alignment with Hierarchical Bottleneck Fusion (DashFusion). Firstly, dual-stream alignment module synchronizes multimodal features through temporal and semantic alignment. Temporal alignment employs cross-modal attention to establish frame-level correspondences among multimodal sequences. Semantic alignment ensures consistency across the feature space through contrastive learning. Secondly, supervised contrastive learning leverages label information to refine the modality features. Finally, hierarchical bottleneck fusion progressively integrates multimodal information through compressed bottleneck tokens, which achieves a balance between performance and computational efficiency. We evaluate DashFusion on three datasets: CMU-MOSI, CMU-MOSEI, and CH-SIMS. Experimental results demonstrate that DashFusion achieves state-of-the-art performance across various metrics, and ablation studies confirm the effectiveness of our alignment and fusion techniques. The codes for our experiments are available at https://github.com/ultramarineX/DashFusion.

  • Open Access

    M3ANet: Multi-scale and Multi-Modal Alignment Network for Brain-Assisted Target Speaker Extraction

    by Cunhang Fan, Ying Chen, Jian Zhou, Zexu Pan, Jingjing Zhang, Youdian Gao, Xiaoke Yang, Zhengqi Wen, Zhao Lv

    The brain-assisted target speaker extraction (TSE) aims to extract the attended speech from mixed speech by utilizing the brain neural activities, for example Electroencephalography (EEG). However, existing models overlook the issue of temporal misalignment between speech and EEG modalities, which hampers TSE performance. In addition, the speech encoder in current models typically uses basic temporal operations (e.g., one-dimensional convolution), which are unable to effectively extract target speaker information. To address these issues, this paper proposes a multi-scale and multi-modal alignment network (M3ANet) for brain-assisted TSE. Specifically, to eliminate the temporal inconsistency between EEG and speech modalities, the modal alignment module that uses a contrastive learning strategy is applied to align the temporal features of both modalities. Additionally, to fully extract speech information, multi-scale convolutions with GroupMamba modules are used as the speech encoder, which scans speech features at each scale from different directions, enabling the model to capture deep sequence information. Experimental results on three publicly available datasets show that the proposed model outperforms current state-of-the-art methods across various evaluation metrics, highlighting the effectiveness of our proposed method. The source code is available at: https://github.com/fchest/M3ANet.


  • Open Access

    M3ANet: Multi-scale and Multi-Modal Alignment Network for Brain-Assisted Target Speaker Extraction

    by Cunhang Fan, Ying Chen, Jian Zhou, Zexu Pan, Jingjing Zhang, Youdian Gao, Xiaoke Yang, Zhengqi Wen, Zhao Lv

    The brain-assisted target speaker extraction (TSE) aims to extract the attended speech from mixed speech by utilizing the brain neural activities, for example Electroencephalography (EEG). However, existing models overlook the issue of temporal misalignment between speech and EEG modalities, which hampers TSE performance. In addition, the speech encoder in current models typically uses basic temporal operations (e.g., one-dimensional convolution), which are unable to effectively extract target speaker information. To address these issues, this paper proposes a multi-scale and multi-modal alignment network (M3ANet) for brain-assisted TSE. Specifically, to eliminate the temporal inconsistency between EEG and speech modalities, the modal alignment module that uses a contrastive learning strategy is applied to align the temporal features of both modalities. Additionally, to fully extract speech information, multi-scale convolutions with GroupMamba modules are used as the speech encoder, which scans speech features at each scale from different directions, enabling the model to capture deep sequence information. Experimental results on three publicly available datasets show that the proposed model outperforms current state-of-the-art methods across various evaluation metrics, highlighting the effectiveness of our proposed method. The source code is available at: https://github.com/fchest/M3ANet.

  • Open Access

    RadialRouter: Structured Representation for Efficient and Robust Large Language Models Routing

    by Ruihan Jin, Pengpeng Shao, Zhengqi Wen, Jinyang Wu, Mingkuan Feng, Shuai Zhang, Jianhua Tao

    Ruihan Jin, Pengpeng Shao, Zhengqi Wen, Jinyang Wu, Mingkuan Feng, Shuai Zhang, Jianhua Tao. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.