Tianyun Zhang

Dr. Tianyun Zhang


Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 October 2018
Verified
Book Chapter
Closed access
PDF file
TZTianyun Zhang
Tianyun Zhang
SYShaokai Ye
Shaokai Ye
KZKaiqi Zhang
Kaiqi Zhang
JTJian Tang
Jian Tang
WWWujie Wen
Wujie Wen
YWYanzhi Wang
Yanzhi Wang

30 August 2023 in Green Energy and Intelligent Transportation
Verified
Journal Article
Open access
XLXinyu Liu
Xinyu Liu
JLJinlong Li
Jinlong Li
JMJin Ma
Jin Ma
HSHuiming Sun
Huiming Sun
ZXZhigang Xu
Zhigang Xu
HYHongkai Yu
Hongkai Yu

15 February 2021 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
TZTianyun Zhang
Tianyun Zhang
SYShaokai Ye
Shaokai Ye
XFXiaoyu Feng
Xiaoyu Feng
XMXiaolong Ma
Xiaolong Ma
KZKaiqi Zhang
Kaiqi Zhang
YWYanzhi Wang
Yanzhi Wang
Weight pruning methods of deep neural networks (DNNs) have been demonstrated to achieve a good model pruning rate without loss of accuracy, thereby alleviating the significant computation/storage requirements of large-scale DNNs. Structured weight pruning methods have been proposed to overcome the limitation of irregular network structure and demonstrated actual GPU acceleration. However, in prior work, the pruning rate (degree of sparsity) and GPU acceleration are limited (to less than 50%) when accuracy needs to be maintained. In this work, we overcome these limitations by proposing a unified, systematic framework of structured weight pruning for DNNs. It is a framework that can be used to induce different types of structured sparsity, such as filterwise, channelwise, and shapewise sparsity, as well as nonstructured sparsity. The proposed framework incorporates stochastic gradient descent (SGD; or ADAM) with alternating direction method of multipliers (ADMM) and can be understood as a dynamic regularization method in which the regularization target is analytically updated in each iteration. Leveraging special characteristics of ADMM, we further propose a progressive, multistep weight pruning framework and a network purification and unused path removal procedure, in order to achieve higher pruning rate without accuracy loss. Without loss of accuracy on the AlexNet model, we achieve $2.58\times $ and $3.65\times $ average measured speedup on two GPUs, clearly outperforming the prior work. The average speedups reach $3.15\times $ and $8.52\times $ when allowing a moderate accuracy loss of 2%. In this case, the model compression for convolutional layers is $15.0\times $ , corresponding to $11.93\times $ measured CPU speedup. As another example, for the ResNet-18 model on the CIFAR-10 data set, we achieve an unprecedented $54.2\times $ structured pruning rate on CONV layers. This is $32\times $ higher pruning rate compared with recent work and can further translate into $7.6\times $ inference time speedup on the Adreno 640 mobile GPU compared with the original, unpruned DNN model. We share our codes and models at the link http://bit.ly/2M0V7DO .

6 May 2020
Verified
Book Chapter
Closed access
PDF file
JLJiayu Li
Jiayu Li
TZTianyun Zhang
Tianyun Zhang
HTHao Tian
Hao Tian
SJShengmin Jin
Shengmin Jin
MFMakan Fardad
Makan Fardad
RZReza Zafarani
Reza Zafarani

13 October 2021 in International Journal of Data Science and Analytics
Verified
Journal Article
Closed access
PDF file
JLJiayu Li
Jiayu Li
TZTianyun Zhang
Tianyun Zhang
HTHao Tian
Hao Tian
SJShengmin Jin
Shengmin Jin
MFMakan Fardad
Makan Fardad
RZReza Zafarani
Reza Zafarani

15 July 2025 in Intelligent Systems with Applications
Verified
Journal Article
Open access
MFMengchen Fan
Mengchen Fan
TZTianyun Zhang
Tianyun Zhang
XMXiaolong Ma
Xiaolong Ma
JGJiacheng Guo
Jiacheng Guo
ZZZheng Zhan
Zheng Zhan
YWYanzhi Wang
Yanzhi Wang