Avatar

Yuan Li

Software Engineer

Advertising Eng & Dev, Amazon

Biography

Hi! This is Yuan Li, New joined Software Engineer and Researcher in Computer Visualization.

My research lines in Data Mining and Computer Visualization. I am curious about explainable and user friendly visualization way for understanding results from Data Mining.

I graduated as Master of Science in Computer Science at New York University. Before that, I received my bachelor degree from School of Software Engineering at Tongji University.

Currently, I am working as Software Engineer at Amazon(NYC).

Feel free to reach me at foxerlee1@gmail.com or yl6606@nyu.edu. Or you can left your message here, I will reply as soon as possible!

Interests

  • Computer Visualization
  • Data Mining
  • Image Segmentation

Education

  • MEng in Computer Science, 2021

    New York University

  • BSc in Software Engineering, 2019

    Tongji University

Skills

Python

Latex

Node.js

Kaggle

Docker

Linux

Experience

 
 
 
 
 

SDE Intern

Advertising Eng & Dev, Amazon

May 2020 – Aug 2020 New York City, United States
Responsibilities include:

  • Built UI for a new React website, replace the original JSP framework with React.
  • Set up service to simulate endpoint calls.
 
 
 
 
 

Graduate Assistant

New York University

Sep 2019 – Feb 2020 New York City, United States
Responsibilities include:

  • Worked at Center for Cybersecurity.
  • Analyzed Relationship between Graphs in Cyber Attack.
  • Built a Factored MDP Approach System to Optimal Mechanism Design for Resilient Large-Scale Interdependent Critical Infrastructures.
 
 
 
 
 

Research Intern

National Instruments

Aug 2018 – Jan 2019 Shanghai, China

Research:

  • Analyzed existing online signal & speech detection and classification algorithms using raw signal data, such as WaveNet, 1D-CNN, and MFCC-CNN.
  • Conceptualized and implemented a more efficient algorithm for fault detection of rotating machinery based on 1D-CNN.
  • Made a great improvement of 2.24% over baseline result leveraging Mel-frequency cepstral coefficients and Convolutional neural network.

Develop:

  • Developed signal energy feature extraction method on Arm Cortex-M using Segger.

  • Wrote unit tests and end-to-end tests to achieved 80% code coverage.

 
 
 
 
 

Software Intern

Microsoft

Jun 2018 – Jul 2018 Shanghai, China
Responsibilities include:

  • Converted Tab files to SQL server format in Azure Data Factory.
  • Assisted the mentor to complete the optimization of the code of iteration 4 in Q2 (second quarter).
  • Analyzed table data from Azure Data Factory and got data distribution by visualization in python.
 
 
 
 
 

Undergraduate Research Assistant

Tongji University

Sep 2016 – Jun 2019 Shanghai, China
Task:

  • Investigated the well-known neural network for Few-shot Learning, Image Segmentation, and Detection.
  • Preprocessed the original CT images.
  • Tested on the dataset provided by the PROSTATEx Challenge, achieving an excellent result compared to other solutions.
  • Built the Neural Network based on U-net, for identifying suspected pulmonary nodules.
  • Published paper《A Hybrid Model: DGnet-SVM for the Classification of Pulmonary Nodules》

Projects

Prostate Cancer Classification for Few-shot Learning

Task:

  • Developed a novel deep learning framework, achieving an excellent result compared to other solutions on the dataset provided by the PROSTATEx Challenge.

  • Outperformed other traditional neural networks with 5.2x speedup and 4.56% acc improvement.

Text Detection of Web Images – ICPR MTWI 2018 Challenge II

Task:

  • Implemented an End-to-End system based on Aster and CTPN.

  • Built an automatic system which can generate Chinese data automatically, containing 10,000 pieces of images, 6-10 chars for each image. Reached 6.72% improvement over baseline.

Intelligent Diagnosis of Pulmonary Nodules

Task:

  • Preprocessed the original CT images.

  • Built the Neural Network based on U-net, for identifying suspected pulmonary nodules.

  • Published paper《A Hybrid Model: DGnet-SVM for the Classification of Pulmonary Nodules》.

Recent Publications

A Hybrid Model: DGnet-SVM for the Classification of Pulmonary Nodules

We investigate the problem of benign and malignant pulmonary nodules classification for thoracic Computed Tomography (CT) images. …

Contact