Avatar

Xingyu Xie

xyxie AT pku.edu.cn

Ph.D. student

Peking University

Personal Site

Biography

I am currently a Ph.D. student at ZERO-LAB, hosted by Prof. Zhouchen Lin. Before joining Peking Unversity, I completed my bachelor’s and master’s degree in 2016 and 2019, respectively, and was fortunate to be co-advised by Prof. Guangcan Liu. My research interests broadly include theories and applications of deep and differentiable models, such as learning-based optimization, the convergence of non-convex/convex optimization, low-rank data recovery, asymptotic analysis, and PDE/ODE.

Interests

  • Deep Learning Theory
  • Non-convex/Convex Optimization
  • Low-Rank Recovery

Education

  • MSc in Computer Vision, 2016-2019

    Nanjing University of Aeronautics and Astronautics

  • BSc in Mechanical Engineering and Automation, 2012-2016

    Nanjing University of Aeronautics and Astronautics

Publications @ZERO Lab

Task-Robust Pre-Training for Worst-Case Downstream Adaptation. NeurIPS, 2023.

Pre-training has achieved remarkable success when transferred to downstream tasks. In machine learning, we care about not only the good …
Jianghui Wang , Yang Chen , Xingyu Xie , Cong Fang , Zhouchen Lin
PDF

Maximum-and-Concatenation Networks. ICML, 2020.

We propose a novel multi-layer DNN structure termed MCN, which can approximate some class of continuous functions arbitrarily well even …
Xingyu Xie , Kong Hao , Jianlong Wu , Wayne Zhang , Guangcan Liu , Zhouchen Lin
PDF

Unified Graph and Low-rank Tensor Learning for Multi-view Clustering. AAAI, 2020.

We propose a novel framework to jointly learn the affinity graph and low-rank tensor decomposition for multi-view clustering.
Jianlong Wu , Xingyu Xie , Liqiang Nie , Zhouchen Lin , Hongbin Zha
PDF

Neural Ordinary Differential Equations with Envolutionary Weights. PRCV, 2019.

Neural networks have been very successful in many learning tasks, for their powerful ability to fit the data. Recently, to understand …
PDF supp

Differentiable Linearized ADMM. ICML, 2019.

We propose D-LADMM, which is a K-layer LADMM inspired deep neural network and rigorously prove that there exist a set of learnable …
PDF

t-Schatten- p Norm for Low-Rank Tensor Recovery. JSTSP, 2018.

We propose a new definition of tensor Schatten-p norm (t-Schatten-p norm) based on t-SVD, and prove that this norm has similar …
PDF

Academic Acativities

Reviewer to Journals

IEEE Transactions on Neural Networks and Learning Systems (TNNLS)
Computer Vision and Image Understanding (CVIU)
Neurocomputing
IEEE Signal Processing Letters (SPL)

Reviewer to Conferences

IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017, 2018
International Conference on Computer Visions (ICCV) 2017
European Conference on Computer Vision(ECCV) 2018
AAAI Conference on Artificial Intelligence (AAAI) 2017, 2018, 2019, 2020
International Joint Conference on Artificial Intelligence (IJCAI) 2017