James Huang
Electrical Engineering graduate with image processing and machine learning skills. Currently seeking a position as a software engineer, to leverage ML knowledge & coding skills to solve problems. I gained cooperation experience of developing algorithms used on defect detection tasks.
Taipei, Taiwan
Professional Experiences
R&D engineer • Compal Electronics.
Jul 2021 - Present
1. Develop a defect detection algorithm with tensorflow, improving the accuracy from 95% up to 98%.
2. Design and improve the process of iteration of the defect detection model, updating per model from weekly to daily.
3. Deploy defect detection algorithm with distributed system using Docekr and k8s, reducing the 30% number of computers used for the algorithm.
Research • NTHU •
Face Enhanced Single Image Super Resolution Based on Facial Priors and Perceptual Loss
Jan 2019 - Oct 2020
1. Use a single model to handle super-resolution tasks on human faces and objects.
2. Consider facial prior knowledge and object texture as loss function.
3. Use training dataset containing human face images and general images to handle most cases
4. Dealing with a single image in real time (0.007s per 256*256 single image)
5. Close objective scores to the other methods which run in real time (PSNR 0.2 more)
EDUCATION HISTORY
Master of Electrical Engineering, Sep. 2018 - Feb. 2021
National Tsing Hua University
Dept. Electrical Engineering
Major: deep learning, computer vision
Master of Electrical Engineering, Sep. 2014 - Jun. 2018
National Tsing Hua University
Dept. Electrical Engineering
SKILL
LANGUAGE
- Image Processing
- Machine Learning
- Deep Learning
- Python
- C++
- Tensorflow
- Pytorch
- Docker
- Mandarin — native
- English — intermediate
