Wenyang Liu

计算机视觉多模态 AI 视觉语言模型

wenyang001@e.ntu.edu.sg 新加坡 CV LinkedIn Scholar

我目前在 Hyundai–NTU–A*STAR Corporate Lab 担任 Research Fellow，与 Assoc. Prof. Wai Kin (Adams) Kong 合作研究。我的研究聚焦于计算机视觉与多模态学习，涵盖图像恢复、视觉识别，以及面向工业感知的基础模型。

我在南洋理工大学获得电气与电子工程博士学位，导师为 Prof. Lap-Pui Chau（IEEE Fellow）和 Assoc. Prof. Kim-Hui Yap，此前在重庆大学获得工学硕士和工学学士学位。以第一作者身份在 CVPR、ACM Multimedia 和 IEEE Transactions on Multimedia 发表论文，累计论文 20 余篇，获得 300 余次引用。

最新消息

2026年10月8日
我们的工业机器人感知论文获内部技术会议 Excellent Paper 认可，并形成专利申请。我们已受邀于 2026 年 10 月 19–21 日，在韩国 Hyundai Uiwang R&D Center 举办的 E-Forest Tech Day 2026 上演示该系统。
2026年7月10日
我们的论文 DriftAD: Visually-Guided Text Drift for Few-Shot Industrial Anomaly Detection 被 ACM International Conference on Multimedia (ACM MM 2026) 接收。这项工作与 Tianyi Liu、Dongshuo Zhang、Kejun Wu 和 Adams Wai-Kin Kong 合作完成。
2026年4月29日
论文被 Information Fusion 接收。
2025年12月1日
论文被 Pattern Recognition 接收。
2025年11月30日
获得南洋理工大学电气与电子工程博士学位。

科研项目

工业机器人视觉感知

2025年11月 – 至今 Hyundai–NTU–A*STAR

基于基础模型开发异常检测、开放词汇感知与 6D 位姿估计方法。搭建抓取流程，在工业零件上验证并完成现场机器人演示。

ACM MM

人机协作

2025年6月 – 2025年11月 Schaeffler–NTU

评测基于 RGB 与深度数据的 3D 人体姿态估计。开发动作重定向方法，将人体运动映射到机器人动作，支持人机协作。

图像水印与安全

2024年12月 – 2025年6月资助方： DSO National Laboratories

开发鲁棒图像水印与无需配对干净图像的自监督水印去除方法。在 JPEG 压缩、噪声和缩放条件下进行性能评测。

JVCIR

开放词汇工业视觉

2024年3月 – 2024年12月 Schaeffler–NTU

开发开放词汇检测与分割方法，识别固定类别之外的物体。研究有限监督条件下的人-物交互识别。

EMNLPACM MMKBS

图像与视频恢复

2021年1月 – 2024年3月资助方：国家研究基金会

开发受损存储中的图像与视频恢复方法。构建轻量级超分辨率模型，以及无需文件元数据的字节–图像融合分类方法。

CVPRNeurIPSIEEE TMM

高效视频与 AI 计算

2017年9月 – 2020年7月重庆大学

通过 H.264 并行解码提升速度、降低能耗，并参与深度学习推理光子加速器研究。

ICPADSDATE

精选论文

DriftAD: Visually-Guided Text Drift for Few-Shot Industrial Anomaly Detection

Wenyang Liu, Tianyi Liu, Dongshuo Zhang,

Proceedings of the ACM International Conference on Multimedia, 2026
ACM MM
CCF A

Abs Bib

We propose DriftAD, a visually-guided CLIP adaptation framework that adapts text representations to local visual context for few-shot industrial anomaly detection, achieving state-of-the-art results across 1-, 2-, and 4-shot settings with up to 1.5-point gains in AUROC and PRO.
@inproceedings{liu2026driftad, url = {https://arxiv.org/abs/2608.23723}, title = {{DriftAD}: Visually-Guided Text Drift for Few-Shot Industrial Anomaly Detection}, author = {Liu, Wenyang and Liu, Tianyi and Zhang, Dongshuo and Wu, Kejun and Kong, Wai Kin (Adams)}, booktitle = {Proceedings of the ACM International Conference on Multimedia}, year = {2026} }
PromptSR: Cascade Prompting for Lightweight Image Super-Resolution

Wenyang Liu, Chen Cai, Jianjun Gao,

IEEE Transactions on Multimedia, 2026
IEEE TMM
CCF A CAS Q1 Top JCR Q1

Abs Bib

We propose PromptSR, a cascade prompting framework for lightweight image super-resolution that leverages prompt-guided restoration to achieve efficient real-world image quality improvement.
@article{liu2026promptsr, url = {https://arxiv.org/abs/2507.04118}, title = {{PromptSR}: Cascade Prompting for Lightweight Image Super-Resolution}, author = {Liu, Wenyang and Cai, Chen and Gao, Jianjun and Wu, Kejun and Wang, Yi and Yap, Kim-Hui and Chau, Lap-Pui}, journal = {IEEE Transactions on Multimedia}, year = {2026}, publisher = {IEEE} }

SSH-Net: A Self-Supervised and Hybrid Network for Noisy Image Watermark Removal

SSH-Net: A Self-Supervised and Hybrid Network for Noisy Image Watermark Removal

Wenyang Liu, Jianjun Gao and Kim-Hui Yap

Journal of Visual Communication and Image Representation, 2025

JVCIR

CCF C CAS Q3 JCR Q2

@article{liu2025ssh,
  url = {https://arxiv.org/abs/2505.05088},
  title = {{SSH-Net}: A Self-Supervised and Hybrid Network for Noisy Image Watermark Removal},
  author = {Liu, Wenyang and Gao, Jianjun and Yap, Kim-Hui},
  journal = {Journal of Visual Communication and Image Representation},
  pages = {104516},
  year = {2025},
  publisher = {Academic Press}
}

ByteNet: Rethinking Multimedia File Fragment Classification through Visual Perspectives

ByteNet: Rethinking Multimedia File Fragment Classification through Visual Perspectives

Wenyang Liu, Kejun Wu, Tianyi Liu,

IEEE Transactions on Multimedia, 2024

IEEE TMM

CCF A CAS Q1 Top JCR Q1

@article{liu2024bytenet,
  url = {https://arxiv.org/abs/2410.20855},
  title = {{ByteNet}: Rethinking Multimedia File Fragment Classification through Visual Perspectives},
  author = {Liu, Wenyang and Wu, Kejun and Liu, Tianyi and Wang, Yi and Yap, Kim-Hui and Chau, Lap-Pui},
  journal = {IEEE Transactions on Multimedia},
  volume = {27},
  pages = {1305--1319},
  year = {2024},
  publisher = {IEEE}
}

Bitstream-Corrupted JPEG Images Are Restorable: Two-Stage Compensation and Alignment Framework for Image Restoration

Bitstream-Corrupted JPEG Images Are Restorable: Two-Stage Compensation and Alignment Framework for Image Restoration

Wenyang Liu, Yi Wang, Kim-Hui Yap,

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023

CVPR

CCF A

@inproceedings{liu2023bitstream_cvpr,
  url = {https://openaccess.thecvf.com/content/CVPR2023/html/Liu_Bitstream-Corrupted_JPEG_Images_Are_Restorable_Two-Stage_Compensation_and_Alignment_Framework_CVPR_2023_paper.html},
  title = {Bitstream-Corrupted {JPEG} Images Are Restorable: Two-Stage Compensation and Alignment Framework for Image Restoration},
  author = {Liu, Wenyang and Wang, Yi and Yap, Kim-Hui and Chau, Lap-Pui},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages = {9979--9988},
  year = {2023}
}

A Byte Sequence Is Worth an Image: CNN for File Fragment Classification Using Bit Shift and n-Gram Embeddings

Wenyang Liu, Yi Wang, Kejun Wu,

IEEE International Conference on Artificial Intelligence Circuits and Systems, 2023
AICAS Oral
Abs Bib

We demonstrate that byte sequences can be treated as images for file fragment classification, using bit shift and n-gram embeddings with CNN architectures.
@inproceedings{liu2023byte, url = {https://arxiv.org/abs/2304.06983}, title = {A Byte Sequence Is Worth an Image: {CNN} for File Fragment Classification Using Bit Shift and n-Gram Embeddings}, author = {Liu, Wenyang and Wang, Yi and Wu, Kejun and Yap, Kim-Hui and Chau, Lap-Pui}, booktitle = {IEEE International Conference on Artificial Intelligence Circuits and Systems}, year = {2023} }

MindReading: An Ultra-Low-Power Photonic Accelerator for EEG-Based Human Intention Recognition

MindReading: An Ultra-Low-Power Photonic Accelerator for EEG-Based Human Intention Recognition

Qian Lou^*, Wenyang Liu^*, Weichen Liu,

* 共同第一作者

Asia and South Pacific Design Automation Conference, 2020

ASP-DAC

CCF C

@inproceedings{lou2020mindreading,
  url = {https://arxiv.org/abs/2001.11954},
  title = {{MindReading}: An Ultra-Low-Power Photonic Accelerator for {EEG}-Based Human Intention Recognition},
  author = {Lou, Qian and Liu, Wenyang and Liu, Weichen and Guo, Feng and Jiang, Lei},
  booktitle = {Asia and South Pacific Design Automation Conference},
  pages = {464--469},
  year = {2020},
  organization = {IEEE}
}

HolyLight: A Nanophotonic Accelerator for Deep Learning in Data Centers

HolyLight: A Nanophotonic Accelerator for Deep Learning in Data Centers

Weichen Liu, Wenyang Liu, Yichen Ye,

本人为第二作者，硕士导师为第一作者

Design, Automation & Test in Europe Conference, 2019

DATE

CCF B 140+ 次引用

@inproceedings{liu2019holylight,
  url = {https://doi.org/10.23919/DATE.2019.8715195},
  title = {{HolyLight}: A Nanophotonic Accelerator for Deep Learning in Data Centers},
  author = {Liu, Weichen and Liu, Wenyang and Ye, Yichen and Lou, Qian and Xie, Yiyuan and Jiang, Lei},
  booktitle = {Design, Automation \& Test in Europe Conference},
  pages = {1483--1488},
  year = {2019},
  organization = {IEEE}
}