關於我

簡介

目前我在華碩的AIOT BU、DIT Dept.工作，基於部門在影像處理的經驗，開發深度學習演算法應用於工業製造改善。

Multi-valued neuron with new learning schemes, Shin-Fu Wu and Shie-Jue Lee, Proceedings of the International Joint Conference on Neural Networks, 2013
Multi-Valued Neuron with Sigmoid Activation Function for Pattern Classification, Shen-Fu Wu, Yu-Shu Chiou and Shie-Jue Lee, AIDM, 2014
Employing local modeling in machine learning based methods for time-series prediction, Shin-Fu Wu and Shie-Jue Lee, Expert Systems with Applications, vol. 42, issue 1, Jan. 2015
工研院南創園區委託計畫：多變數時間序列預測模型
Master Thesis: A Novel Forecasting Method for Time Series Data with Missing Values

Prognostic Health Management: MOCVD - particle filter (next run failure/remaining useful life) prediction
Virtual Metrology: milling tool chipping detection

Cupper Plating FDC: Monitoring machine status by its PLC and IOT signals.
Production Time Estimation: Estimate the production time by corresponding coefficients in run card.
GUI Visualization: Building windows softwares by PyQt5 in Python.
Motor PHM: Building fault detection mechanism using motor current & vibration.
Gerber Text Matching: Transforming the text in a gerber file to vector signatures for text recognition.
Circuit Auto-Routing: Seeking inspiration from path-finding algorithms in game industry and computer graphics. Still in a long way to go!
Substrate Design CAD: Parameterize design specs for DXF-file auto generation.

Study:
- Audio Signal Separation: Paper reading and using Open-Unmix for signal separation
- EasyTL: Transfer Learning paper reading and implement EasyTL in Python(pyEasyTL)
- Heart Sound Classification Research: Using CINC2016 AISTATS2012 datasets for heart sound classification research
Projects:
- MB connector classification demo
- TinkerEdgeT object detection demo
- ABB robot with camera on NB defect detection demo (socker communication with ABB robot)
- AISVision:
  - Classification model develop
  - Classification and Object Detection Model Owner
  - Oriented Object Detection Owner
Competitions:
- #1 on ASUS RoboACE 2022 competition (Robot Football Tournament in ASUS, Team: DIT一拳超人)
- #3 on 2022 Intel Dev Cup competition in “AI in Experiment track” (Team: CyberAudio 2022)
- 2024 AICUP AI 驅動出行未來：跨相機多目標車輛追蹤競賽－模型組 (TEAM_5045 企業特別獎，private board #2)

Python: 工作後幾乎都在用Python，已經快變成我的母語了。
C/C++: 大學第一個程式語言、嵌入式系統也有用C寫過，研究所實作影像處理用C++。目前工作是在撰寫跨語言(C/C++/C#/Python)的DLL時候使用。
Matlab: 整個研究所的研究成果都是用Matlab跑的，不過現在叫我做資料分析應該也是果斷跳槽用Python。
Java: 大學修演算法的時候用過，退化到幾乎不知道的程度。之後在工研院上課也有用到，台大資工老師的課滿硬的，加減學了一下。