Philip Wootaek Shin

Have experience working in Industry for 3 years. Specification: Machine Learning, Computer Vision

Dolby Laboratories Sunnyvale, CA

Imaging Research Intern (Advanced Technology Group) May 2023 – August 2023

• Topic: Floor Plan Construction for Multiple Perspective Video using Object Based Latent Vector Aggregation
  • Specifics: User Generate Content (UGC) floor plan retrieval.
  • Manager: Guan-Ming Su

Research Engineer (OLED TV Operating Circuit Design Team/Vision Algorithm Task) Jul. 2021 – Present

• Upgrade Deep-Learning-based Automatic Defect Object Detection Model Development using OLED Panel Images.
  • Main Issues: Tiny Defect Detection, Small Dataset Training due to Periodic Panel Modification, and Data Imbalance.
  • Improve the current detection rate by 10% and optimize data augmentation for detection system.
• Design Deep-Learning based Automatic Classification Model to Defect TFT Panel Images.
  • Main Issues: Edge Inference using CPU, Huge image size with small size defects, and Data Imbalance.
  • Enhance a commercialized in-usage tool by 5%; crop and resize the image for improvement in classification.
• Devise a Pilot Defect Recognition Tool based on Super Resolution.
  • Main Issues: Low-resolution image for human eye and gray scale level indistinguishable for human eye.
  • Develop a deep learning pilot system to automatically convert low-resolution indistinguishable image to high resolution.
• Built Anomaly Detection Automatic Email System that sends detection of stains.
  • Main Issues: Indistinguishable Stain Detection, Border Line Detection, and Database management. 
  • Constructed a Real-time Anomaly Detection System from the pilot auto-email system; maintained the current email system. 
• Develop a Multi-anomaly Detection Tool for user-friendly interface to train and test different models.
  • Main Issues: Deployment of different anomaly detection strategy in one interface, optimization of each method
  • Designed a GUI interface for in-house usage and differentiated individual anomaly detection methods.

Assistant Research Engineer (Data Governance Team/AI Part Leader) July 2019 – June 2021

• Built a Term/Dictionary Recommendation System using Machine Learning and Natural Language Processing.
  • Main Issues: Small Dataset due to NDA with Client; needed both English and Korean recommendation systems.
  • Implemented API to connect the technical glossary and the recommendation system. 
  • Designed a UI/UX interface for top 5 terms recommendation for terms and domain in the glossary.
• Designed and programmed a Machine Learning-based Catalog System for Recommendation.
  • Main Issues: No state-of-the-art recommendation tool and difficulty in measuring the accuracy of recommendation.
  • Implemented 4 different recommendation techniques for data catalogs.
  • Constructed a database-based user tracking system for user’s personalized data and recommendation.
• Developed and implemented data virtualization commercialized tool with various DBMS support
  • Main Issues: Expansion of different DBMS and optimization, and customization for clients.
  • Implemented a Hadoop-based virtualization tool and a Query repository for handling and managing different data.
  • Experimented an AI-based caching algorithm and scheduling method for the constructed system.
• Achievements: 4 publications, 3 Korean patent registrations, 1 U.S. patent application; August 2020 Employee of the Month Award

EDUCATION The Pennsylvania State University University Park, PA

Doctor of Philosophy in Computer Science and Engineering. Aug. 2023 – Present

• Advisor: Vijaykrishnan Narayanan, Jack Sampson

Doctor of Philosophy in Electrical and Computer Engineering. Aug. 2022 – May 2023

• UT Graduate Excellence Fellowship (2022 Fall, 2023 Spring). 
• Temple Foundation Graduate MCD Engineering Fellowship (2022- 2025)

Pennsylvania State University University Park, PA

Master of Science in Computer Science and Engineering. Aug. 2017 – May 2019

• Master’s Thesis: Context Aware Collaborative Object Recognition for Multi Camera Time Series Data

• Bachelor’s Thesis: Coupled Oscillator-based FAST Corner Detection
• Graduated with a Magna Cum Laude
• Schreyer Honors College; Dean’s List for All Semesters