Kun FANG

About

My research interests focus on trustworthy AI, mainly on the adversarial and certified robustness of deep neural networks, and the topic of out-of-distribution detection. I also have previously dabbled in the kernel methods in machine learning.

Work

Hong Kong Polytechnic University
|

Postdoctoral Fellow

Hong Kong

→

Education

Shanghai Jiao Tong University
China

→

Ph.D.

Shanghai Jiao Tong University
China

→

Master

Tongji University
China

→

Bachelor

Publications

Multi-head ensemble of smoothed classifiers for certified robustness

Published by

Neural Networks

Summary

journal-article

Revisiting Deep Ensemble for Out-of-Distribution Detection: A Loss Landscape Perspective

Published by

International Journal of Computer Vision

Summary

journal-article

Towards robust neural networks via orthogonal diversity

Published by

Pattern Recognition

Summary

journal-article

Kernel PCA for Out-of-Distribution Detection

Published by

Advances in Neural Information Processing Systems

Summary

conference-paper

Unifying Gradients to Improve Real-World Robustness for Deep Networks

Published by

ACM Transactions on Intelligent Systems and Technology

Summary

journal-article

Improving the adversarial robustness of quantized neural networks via exploiting the feature diversity

Published by

Pattern Recognition Letters

Summary

journal-article

Improving adversarial robustness through a curriculum-guided reliable distillation

Published by

Computers & Security

Summary

journal-article

Local-Global Semantic Fusion Single-shot Classification Method

Published by

International Conference on Neural Information Processing

Summary

conference-paper

End-to-end kernel learning via generative random Fourier features

Published by

Pattern Recognition

Summary

journal-article

CROON: Automatic Multi-LiDAR Calibration and Refinement Method in Road Scene

Published by

2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Summary

conference-paper

Subspace Adversarial Training

Published by

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Summary

conference-paper

ADVMIX: Data Augmentation for Accurate Scene Text Spotting

Published by

2021 IEEE International Conference on Image Processing (ICIP)

Summary

conference-paper

Towards Unbiased Random Features with Lower Variance For Stationary Indefinite Kernels

Published by

2021 International Joint Conference on Neural Networks (IJCNN)

Summary

conference-paper