Publications | Luchen Zhou

2026 DAC CCF-A Published

TRIDENT: An End-to-End Streaming Accelerator for TriSpGEMM

Yiyue Hu, An Hu, Hui Guo, Luchen Zhou, Yongzhang Nie, Runzhang Mao, Gaoyang Zhao, Xia Zhao, Yongwen Wang

This work studies an end-to-end streaming accelerator for TriSpGEMM and is now published at DAC. It broadens the site's research profile toward high-performance systems and accelerator design.

Accelerator Streaming Systems CCF-A

2026 Architecture Venues CCF-A Revised Manuscript

GRACE-Infer: A Stage-Aware Multi-Granularity Framework with Metadata Reuse for Transformer Inference

Luchen Zhou, Hao Sun, Yiyue Hu, Ziming Chen, Yongzhang Nie, Yongwen Wang

Proposes a stage-aware multi-granularity framework for transformer inference with metadata reuse. The latest version is positioned as a revised manuscript targeting CCF-A computer architecture venues, emphasizing systems efficiency, reuse-aware execution, and scalable inference design.

Transformer Inference Metadata Reuse CCF-A

2025 Computers & Security CCF-B Published

E-WebGuard: Enhanced Neural Architectures for Precision Web Attack Detection

Luchen Zhou, Wei-Chuen Yau, Y.S. Gan, Sze-Teng Liong

Presents enhanced neural architectures for more precise web attack detection. The paper demonstrates strong performance improvements over earlier deep learning baselines and helps connect learning-based security models with deployable classification systems.

Security AI Web Detection Neural Models

2024 ESWA CCF-B Published

DSteganoM: Deep Steganography for Motion Capture Data

Qi Wen Gan, Wei-Chuen Yau, Y.S. Gan, Iftekhar Salam, Shihui Guo, Chin-Chen Chang, Yubing Wu, Luchen Zhou

Explores deep steganography in motion capture data, expanding the research portfolio toward data representation, encoding, and applied machine learning settings.

Representation Steganography Applied AI

2022 IEEE CCET EI Published

Detecting Web Application Injection Attacks Using One-Class SVM

Luchen Zhou, Tao Lu, Xiaobo Hu

Investigates an anomaly-detection approach to web application injection attacks using one-class SVM, with a 94.04% detection rate and 1.62% false positive rate reported in the current clean resume.

One-Class SVM Injection Attacks Anomaly Detection

2024 Frontiers in Marine Science Published

Nodal Modulation of M2 and N2 Tides along the Norwegian Coast

Zong, X., Zhou, J., Yang, M., Zhang, S., Deng, F., Lian, Q., ... & Chen, Z.

Appears in the latest clean resume as an additional co-authored publication, extending the publication list beyond AI systems and security into interdisciplinary scientific research.