Peiyu Yang

I am a researcher with interests in explainable artificial intelligence and trustworthy machine learning. I am currently a Research Fellow at School of Computing and Information Systems, The University of Melbourne, advised by Dr Feng Liu, Prof Sarah Erfani , and Prof Christopher Leckie . I finished my PhD from The University of Western Australia, advised by Dr Naveed Akhtar, Dr Zeyi Wen and Prof Ajmal Mian.

Email  /  Google Scholar  /  Github

profile photo
News

Feb., 2026: Our paper Attribution-Guided Model Rectification of Unreliable Neural Network Behaviors has been accpeted by CVPR 2026!

Dec., 2024: I graduated with a PhD from The University of Western Australia! 🎓 🎉

Jul., 2024: Our paper Regulating Model Reliance on Non-Robust Features by Smoothing Input Marginal Density has been accpeted by ECCV 2024!

Jan., 2023: Our paper Re-Calibrating Feature Attributions for Model Interpretation has been accpeted by ICLR 2023 Spotlight!

Dec., 2022: Our paper Local Path Integration for Attribution has been accpeted by AAAI 2023!

Publications

     † indicates equal contribution. * indicates corresponding author.

prl Towards Efficient Large Language Model Serving: A Survey on System-Aware KV Cache Optimization
Jiantong Jiang, Peiyu Yang*, Rui Zhang, Feng Liu.
Under Review.

Code/bibtex
prl Dynamic Model Editing to Rectify Unreliable Behavior in Neural Networks
Peiyu Yang, Naveed Akhtar, Jiantong Jiang, Ajmal Mian.
Computer Vision and Pattern Recognition (CVPR), 2026.

Code/bibtex
prl Backdoor-based Explainable AI Benchmark for High Fidelity Evaluation of Attribution Methods
Peiyu Yang, Naveed Akhtar, Jiantong Jiang, Ajmal Mian.
Under Review.

Code/bibtex
prl Fast-PGM: Fast Probabilistic Graphical Model Learning and Inference
Jiantong Jiang, Zeyi Wen, Peiyu Yang, Atif Mansoor, Ajmal Mian.
Under Review.

Code/bibtex
prl Regulating Model Reliance on Non-Robust Features by Smoothing Input Marginal Density
Peiyu Yang, Naveed Akhtar, Mubarak Shah, Ajmal Mian.
European Conference on Computer Vision (ECCV), 2024.

Code/bibtex
prl Re-Calibrating Feature Attributions for Model Interpretation
Peiyu Yang, Naveed Akhtar, Zeyi Wen, Mubarak Shah, Ajmal Mian.
International Conference on Learning Representations (ICLR Spotlight), 2023.

Code/bibtex
prl Local Path Integration for Attribution
Peiyu Yang, Naveed Akhtar, Zeyi Wen, Ajmal Mian.
AAAI Conference on Artificial Intelligence (AAAI), 2023.

Code/bibtex
prl Multi-Grained Interpretable Network for Image Recognition
Peiyu Yang, Zeyi Wen, Ajmal Mian.
International Conference on Pattern Recognition (ICPR), 2022.

bibtex
prl Visible Part Prediction and Temporal Calibration for Pedestrian Detection
Peiyu Yang, Weixi Li, Lu Wang, Lisheng Xu, Qingxu Deng.
IET Image Processing (IETIP), 2023.

bibtex
prl A Part-Aware Multi-Scale Fully Convolutional Network for Pedestrian Detection
Peiyu Yang, Guofeng Zhang, Lu Wang, Lisheng Xu, Qingxu Deng, Ming-Hsuan Yang.
IEEE Transaction on on Intelligent Transportation Systems (TITS), 2021.

Code/bibtex
prl Hardhat-Wearing Detection Based on a Lightweight Convolutional Neural Network with Multi-Scale Features and a Top-Down Module
Lu Wang, Liangbin Xie, Peiyu Yang, Qingxu Deng, Shuo Du, Lisheng Xu.
Sensor, 2022.

bibtex
prl A Local Top-Down Module for Object Detection with Multi-scale Features
Shihua Huang, Lu Wang, Peiyu Yang, Qingxu Deng.
Chinese Conference on Pattern Recognition and Computer Vision (PRCV), 2022.

bibtex
Professional Service

Reviewer for ICLR 2025, NeurIPS 2024, ICML 2024-2025, CVPR 2024-2025, ECCV 2024, etc.