Wan Wei Fan

Senior Engineer

/Algorithm Engineer

/Data Scientist

[email protected] 

+886 0933-974-183

Hsinchu ,  Taiwan

Love to absorb new knowledge and try technologies in different fields. Focus on the development and application of algorithms in different areas, and be familiar with various development tools. Have experience of building edge computing systems from scratch, and successfully applied it to more than three fields and multiple deployments. Mainly working in the development of AI Algorithms Application/Machine Vision/Robot Vision.

Skills

Programming Languages


  • Python
  • Matlab
  • C/C++

Packages


  • OpenCV
  • TensorFlow/Keras
  • Pytorch
  • pandas, scikit-learn

Knowledge


  • Deep Learning, Data Analysis
  • Computer Vision
  • Image Processing
  • Robot Vision

Other Tools


  • VS Code, Jupyter, Spyder
  • Docker, Git, Anaconda
  • Nvidia Jetson series embedded system
  • Linux/Ubuntu, Taiwan Computing Cloud, ROS/ROS2
  • 3D camera application (Intel Realsense series, Microsoft Azure Kinect)

Work Experience

May 2020 - Now

Senior Engineer/Algorithm Engineer/Data Scientist

AU Optronics Corp.

  • AI and vision algorithm research and development.
  • Robot vision algorithm research and development.
  • Camera/hand-eye calibration and 3D camera applications.
  • AI Embedded system product development.

2017 - 2019

Automatic Control TA

National Chung Hsing University

  • Teaching students to use MATLAB/Simulink.
  • Scoring the test and correcting errors.

2017 - 2018

Numerical Analysis TA

National Chung Hsing University

  • Numerical analysis method teaching.
  • Reviewing homework on Matlab.

Education

2017 - 2020

National Chung Hsing University

Department of Mechanical Engineering

Intelligent Control and Applications Laboratory (now moved to National Yang Ming Chiao Tung University)

2013 - 2017

National Central University

Department of Mechanical Engineering

Group of Opto-Mechatronics

Projects


Hand-eye Calibration Flow

  • Implemented by OpenCV and ROS2.
  • Automatically test the accuracy of the calibration results.
  • Automatically detect the correctness of the calibration data.

Azure Kinect Yolov5 3D Positioner

  • Showing the relative 3D position of the detected object and the camera.
  • Implemented by OpenCV and pytorch.
  • Azure Kinect data is read in python.

Vision Assistant

  • Used to ensure workflow and safety.
  • Multiple AI models and framework applications.
  • Connecting different machines and systems.
  • System log save and upload.


Two Stages Unsupervised Object Detection

  • Object Detection only uses class labels.
  • Technology presentation demo at ICA Lab.
  • A combination of unsupervised learning and image processing methods.
  • Implemented by TensorFlow/Keras and OpenCV.
  • Applied to two types of data set of defect.

Master Thesis

  • A Convolutional Neural Network with Adding Feature Space Noise for Imbalanced Data Classification.
  • Improve CNN performance by noise adding method.
  • Three different data sets are used to test the theory.

Other Experience

13th International Automatic Control Conference (CACS 2019) 

Presenting the research results of defect detection and deep learning in English.

2019 IMBD

Team up with laboratory classmates to participate in the competition and enter the finals.

Powered by CakeResumePowered by CakeResume