Publications

*: equal contribution

2025

[ICML 2025] Velibor Bojković*, Xiaofeng Wu*, et al. Temporal Misalignment in ANN-SNN Conversion and Its Mitigation via Probabilistic Spiking Neurons. In International Conference on Machine Learning. 2025. [arXiv] [OpenReview] [code] [material] [poster]

Lay Summary

Spiking neural networks (SNNs) are designed to mimic biological functioning of biological neurons. In this work we discover and study a seemingly counterintuitive phenomenon in ANN-to-SNN conversion (a way to train SNNs via pretrained Artificial Neural Networks), which we term Temporal Misalignment. Namely, we find that random permutations of spike trains after spiking layers significantly boost model performance. We dive deeper into explaining what is actually happening and we introduce the Two-Phase Probabilistic (TPP) Spiking Neuron—a novel, biologically plausible, and hardware-friendly neuron model. Mitigating Temporal Misalignment, our TPP neuron enables SNNs to closely match ANN accuracy with just a few time steps.

2024

[ECCV 2024] Xiaofeng Wu*, Velibor Bojković*, Bin Gu, Kun Suo, Kai Zou. FTBC: Forward Temporal Bias Correction for Optimizing ANN-SNN Conversion. In European Conference on Computer Vision. 2024. [paper] [arXiv] [code] [poster]

[ACM SE 2024] Kun Suo, Long Vu, Md. Romyull Islam, Nobel Dhar, Tu N. Nguyen, Selena He, Xiaofeng Wu. A Systematic Investigation of Hardware and Software in Electric Vehicular Platform. In ACM Southeast Conference. 2024. [paper] [slides]

[ACM SE 2024] Nobel Dhar, Bobin Deng, Dan C. Lo, Xiaofeng Wu, Liang Zhao, Kun Suo. An Empirical Analysis and Resource Footprint Study of Deploying Large Language Models on Edge Devices. In ACM Southeast Conference. 2024. [paper] [slides]

2021

[Middleware 2021] Xiaofeng Wu, Jia Rao, Wei Chen, Hang Huang, Chris H. Q. Ding, Heng Huang. SwitchFlow: Preemptive Multitasking for Deep Learning. In International Middleware Conference. 2021. [Best Paper Award] [Award PDF] (1/107 submissions) [paper] [slides] [video] [code]

[HPDC 2021] Hang Huang, Jia Rao, Song Wu, Hai Jin, Hong Jiang, Hao Che, Xiaofeng Wu. Towards Exploiting CPU Elasticity via Efficient Thread Oversubscription. In International Symposium on High-Performance Parallel and Distributed Computing. 2021. [paper]

2019

[HPDC 2019] Yong Zhao, Kun Suo, Xiaofeng Wu, Jia Rao, Song Wu, Hai Jin. Preemptive Multi-Queue Fair Queuing. In International Symposium on High-Performance Parallel and Distributed Computing. 2019. [paper]

[HPDC 2019] Hang Huang, Jia Rao, Song Wu, Hai Jin, Kun Suo, Xiaofeng Wu. Adaptive Resource Views for Containers. In International Symposium on High-Performance Parallel and Distributed Computing. 2019. [paper]

2018

[HotCloud 2018] Xiaofeng Wu, Kun Suo, Yong Zhao, Jia Rao. A Side-channel Attack on HotSpot Heap Management. In USENIX Workshop on Hot Topics in Cloud Computing. 2018. [paper] [slides]