Researching on binary semantic segmentation for small objects
Irvine, CA, US
01/2024 ~ 02/2024
Global SW/AI Internship Program
Irvine Tech Hub
Develop modules that can automatically process company-specific documents such as invoices, sales orders, and time sheets through OCR, LLM, and few-shot learning and apply them to systems such as ERPNext.
Projects
Daejeon, Republic of Korea
05/2022 ~ 05/2024
Microplastic Segmentation/Counting (in sewer)
DIP Challenging Lab
Chungnam National University
Detection system for microplastics filtered through wire mesh from sewage - Researched on microplastic detection models based on semantic segmentation and object detection - Tensorflow
05/2023 ~ 03/2024
Daejeon, Republic of Korea
Drunken Driver Detection
DIP Challenging Lab
Chungnam National University
A system that detects drunk drivers through CCTV footage to catch them in advance - Developed a 3D Resnet-18-based model to determine the probability of a person's drunken actions from video input - Tensorflow
12/2021 ~ 03/2023
Daejeon, Republic of Korea
PM Avoidance Application
DIP Challenging Lab
Chungnam National University
Object detection-based rear object detection notification system that attaches to personal mobile devices - Deviced structural design for real-time risk detection display even in LTE environment - Implemented overall screen composition of the app - Android