Avatar of 劉瀚文.
劉瀚文
HannStar AI Engineer
個人檔案
職場能力評價0

貼文
0個聯絡人
列印
Avatar of the user.

劉瀚文

HannStar AI Engineer
Graduated from the Master's program in Computer Science and Information Engineering at National Taiwan Normal University. During my graduate studies, focused on researching topics within the AIoT and embedded systems. Communication, low power consumption, and limited resources are prominent topics in this field. Now I hold positions as an Artificial Intelligence Engineer and Data Scientist at HannStar Display Corporation.
Logo of the organization.
HannStar Display Corp.
Logo of the organization.
National Taiwan Normal University
台北市, 台灣

專業背景

  • 目前狀態
    就職中
    目前會考慮了解新的機會
  • 專業
    數據科學家
    機器學習工程師
    BI 開發人員
  • 產業
    硬體
  • 工作年資
    2 到 4 年 (2 到 4 年相關工作經驗)
  • 管理經歷
  • 技能
    Python
    C++ and C
    AI
    Git
  • 語言能力
    Chinese
    母語或雙語
    English
    進階
  • 最高學歷
    碩士

求職偏好

  • 預期工作模式
    全職
    對遠端工作有興趣
  • 希望獲得的職位
    AI工程師、機器學習工程師、深度學習工程師、資料科學家、Machine Learning Engineer、Deep Learning Engineer、Data Scientist
  • 期望的工作地點
    台北市, 台灣
    台南市, 台灣
    新竹市, 台灣
  • 接案服務

工作經驗

Logo of the organization.

Artificial Intelligence Engineer

2023年5月 - 現在
台北市, 台灣
Main Work Objective: Digital Transformation for Factories Implement AI-assisted detection standards for factory safety management units. Develop auxiliary tools to enhance the efficiency of various units. Responsible for data integration of factory manufacturing machines, introducing automated data updates to enhance factory operational efficiency. Projects Executed: Factory Safety Project: Developed a factory safety alarm system: Responsible for data labeling, fine-tuning, and evaluating the YOLOv5 model, and algorithm development. Using pre-trained models reduces training costs and improves detection performance. This system can automatically detect whether workers comply with factory safety regulations and send alerts to factory managers. With a sample size of 10,000, the correct detection rate reached 82.6%. Tools Used: Python, PyTorch, IPcam Company Text Comparison System: Used OCR text recognition models to capture text from videos or images, ensuring the accuracy of text in different versions of promotional media. Applied across company departments, such as English-Chinese text comparison in promotional videos. Tools Used: Python, Tesserocr AI Integration of Company Portal and Internal Website: Integrated internal chat assistants with the internal website and Dialogflow to provide efficient internal website navigation services, such as HR system guidance, and added value-added service systems. 3-1. Value-added Service Systems: Business Card Scanning System: Extracts text from business card images using OCR models and uses regular expressions to extract personal information, automatically adding it to the contact list. Customized Web Crawler: Based on business needs, customizes the web crawling process to extract detailed information about target companies. Users only need to provide the target company's name or registration number to retrieve information such as the person in charge, capital amount, and board members. Tools Used: Python, Tesserocr, OpenCV, Selenium, BeautifulSoup Visualization Report System: Collected and cleaned data from multiple offline production machines, integrated it, and uploaded it to the database to generate daily, weekly, and monthly reports as needed. Created graphical reports using Tableau and embedded them into the company's internal system. Tools Used: Python, Pandas, Tableau
Logo of the organization.

System Integration testing engineer

2020年3月 - 2021年2月
1 年 0 個月
桃園市, 台灣
1. Testing Compatibility and Stability of Various Server Products in Development and Sales: 1-1. Compatibility and durability testing of various hardware combinations (CPU, MEM, POWER SUPPLY, DISK, PCIe devices, etc.). 1-2. Compatibility and qualitative testing of various operating system versions. 1-3. Assessing heat dissipation and component interaction under high load conditions. 1-4. BMC functionality testing, and testing the compatibility of BMC firmware versions with BIOS versions. 2. Testing Pre-shipment Products for Operating System Certification: 2-1. Performing certification tests according to the specifications of each operating system to obtain official certification. 2-2. Collecting logs from certification tests of various hardware combinations to assist the RD team in the development of next-generation products. 3. Participating in the Vendor Debugging Process: 3-1. Collecting and reproducing all logs when known issues occur and reporting them to the development team in India. 3-2. Setting up reproducible environments for remote monitoring by the original developer in India to assist in debugging. 4. Testing Company-Developed Automated Testing Software for Full Automation of Machine Testing: 4-1. Setting up automated testing environments (both hardware and software). 4-2. Using automated test scripts to test machines and collect logs to ensure the completeness and accuracy of the testing process. 4-3. Collaborating with the RD team to improve automated scripts and enhance the efficiency of automated testing.

學歷

Logo of the organization.
碩士學位
Computer and Information Sciences, General
2021 - 2023
簡介
Course Projects: Real-time Systems: Multi-thread collaboration strategy (Pthread) Rate Monotonic Scheduling implementation Socket implementation in C++, Python, Java Neural Networks: Developed a stock price prediction model using CNN (TensorFlow, Keras) Created an automotive image GAN model (TensorFlow, Keras) Group project: Scene text localization and detection model (based on YOLO detection model) IoT Applications: Facial emotion recognition music recommendation system (Raspberry Pi, Arduino UNO) Data Visualization: Dynamic comparison chart of baseball player performance Group project: Analysis of traffic accident types and casualties in Taipei City Master's Thesis: Smart-City Collaborative Collision Warning for Blocked-View Corners Implemented vehicle position tracking on a lightweight edge platform (Nvidia Jetson Nano) using YOLOv5 and DeepSort for vehicle speed detection and prediction, providing early warnings to other road users. (PyTorch, TensorRT, LoRa, GCP)

職場能力評價