Shangxi Wu 吴尚锡

  Beijing Jiaotong University, Beijing, China
CCF member, IEEE SSA, ACM member

  kirinng0709@gmail.com

Beijing Jiaotong University, Computer Science, July 2016-Current.
My research interests are in computer vision and deep learning.

GPA: 3.93

Publications

link to: Google Scholar Profile
link to: dblp

   

An Experimental Study of Semantic Continuity for Deep Learning Models
Shangxi Wu, Jitao Sang, Xian Zhao, Lizhang Chen
Tech report, arXiv, 2020
arXiv 2011.09789

   

Adaptive Adversarial Logits Pairing
Shangxi Wu, Jitao Sang, Kaiyuan Xu, Guanhua Zheng, Changsheng Xu
Tech report, arXiv, 2020
arXiv 2005.11904

   

Attention, Please! Adversarial Defense via Attention Rectification and Preservation
Shangxi Wu, Jitao Sang, Kaiyuan Xu, Jiaming Zhang, Yanfeng Sun, Liping Jing, Jian Yu.
Tech report, arXiv, 2018
arXiv 1811.09831

   

Robust CAPTCHAs towards Malicious OCR (IEEE Trans MM, 2020)
Jiaming Zhang, Jitao Sang, Kaiyuan Xu, Shangxi Wu, Xian Zhao, Yongli Hu, Yanfeng Sun, Jian Yu.
Tech report, arXiv, 2019
arXiv 1904.09804

   

Butterfly Detection and Classification Based on Integrated YOLO Algorithm (ICGEC, 2019)
Bohan Liang, Shangxi Wu, Kaiyuan Xu, Jingyu Hao.
Tech report, arXiv, 2020
arXiv 2001.00361

   

Image Classification Method Based on Attention Mechanism to Defend Adversarial Examples (Patent)
Shangxi Wu, Jitao Sang, Kaiyuan Xu, Jiaming Zhang, Yanfeng Sun, Liping Jing, Jian Yu.
First inventor of patent

   

CAPTCHA Generation Method of Defensive Character Recognition System Based on Adversarial Examples (Patent)
Jiaming Zhang, Jitao Sang, Kaiyuan Xu, Shangxi Wu, Yanfeng Sun, Yongli Hu, Jian Yu.
Fourth inventor of patent

Research Experience

   

PCL, Shenzhen
Oct. 2019 – Current
Augmentation & small data

   

Beijing Key Lab of Traffic Data Analysis and Mining
Nov. 2017 – Current

   

Institute of Network Science and Intelligent System, Beijing Key Laboratory
Jul. 2017 – Jun. 2018

   

Butterfly Detection Competition, China Conference on Data Mining
Sep. 2017 – Jun. 2018
Accurately identified more than 100 existing Chinese butterflies, get the 13th in classification task and the 20th in image location task (358 teams in total)

SKILLS

   

Programming languages: Python3(Proficient), C, ,OpenCV, Latex
Deep learning framework: Tensorflow(Proficient)