Jun Chuan Chiew

Jun Chuan Chiew

Software Engineer & Computer Vision Enthusiast

Hi! I’m Jun Chuan Chiew. I was born in Malaysia and I currently live in Taipei.

Currently, I’m working at ASUS-AICS as a Software Engineer. Before that, I have worked on several products which leverage Computer Vision to do industrial application, such as smart checkout, survelliance.

I believe that AI software productization can be beneficial to the society / world and I hope I can make contribution in this field.

Interests
  • Deep Learning
  • Computer Vision
  • Machine Learning Operations
  • Cloud Native Applications
Education

  • BSc in Risk Management and Insurance, 2012

    National Chengchi University

Experience

 
 
 
 
 
ASUS-AICS
Senior Software Engineer
ASUS-AICS
May 2020 – Present Taipei
  • Developed cross device tracking model, which achived 0.75 F1-score, and data pipeline to process and inference 20 millions of web logs within 5 hours per day, by using Azure Databricks and Apache Spark.
 
 
 
 
 
Umbo Computer Vision
AI Engineer
Umbo Computer Vision
Aug 2019 – May 2020 Taipei
  • Developed deep learning training pipeline with Kubeflow Pipeline, GKE & TWCC, to scale up capacity and efficiency of model production.
  • Studied and evaluated false alarm filter algorithm, to improve precision of product and to support over ten thousand events.
  • Maintained CV services & tracking business application (Tailgating), including builed monitoring system pipeline, to support thousand of camera streams.
 
 
 
 
 
Viscovery
Computer Vision Engineer
Viscovery
May 2018 – Aug 2019 Taipei
  • Developed smart-checkout system with the algorithms including Metric Learning, Object Detection & Segmentation, to more than 5 clients and more than 5 demo exhibition, with over 0.9 accuracy.
  • Leaded a 2-person team to work on Bread Recognition algorithms, such as bread’s topping augmentation, hierarchical & fine-grained classification and instance segmentation.
  • Implemented cut-paste based data generation tools with tranditional computer vision algorithms, including contour extraction, data augmentation, bluring and color space processing, to increase quantity and variety of training data, while reduce cost of data collection.
  • Studied, and evaluated new and the cutting-edge method, especially Generative Model, to improve feature representation and performance of new products recognition in smart-checkout system without retraining model.
 
 
 
 
 
Academia Sinica - Data Insights Research Lab
Research Assistant
Academia Sinica - Data Insights Research Lab
Jul 2016 – Apr 2018 Taipei
  • Implemented more than 10 Machine Learning Projects to clients, for example, using text mining and XGBoost to model book sales prediction with 0.77 F1-score and applying multi-label classification and deep neural network to model dye selection and optimization with 0.99 Top-10 Accuracy.
  • Leaded a 5-person team to work on Governance satisfaction analysis with App’s data, such as apps logs preprocessing, text mining, data analysis, regression model.
  • Preprocessed and analyzed more than 100 millions / 100GB scales of data, for example, e-commerce’s transaction logs, and applied Apriori algorithm on it to figure out which products or categories combination was the best seller.
  • Taught more than 200 students in the courses of Deep Neural Network, Convolutional Neural Network, Natural Language Preprocessing, and assisted them to work on Machine Learning Projects, for example, using text mining to model artical classification and applying XGBoost and LSTM to model stock price prediction.
  • Consulted several departments and companies to define Machine Learning application fields.

Contact