Hyojeong Park

Soongsil University

School of Electronic Engineering · Seoul, Korea

B.S. (in progress)

Expected graduation: Feb. 2029

GPA: 3.91 / 4.5

ARM Instruction-Based Image Processing Optimization

Soongsil University · Microprocessor Applications Course Project

• Contribution: Participated in comparing and analyzing execution performance across memory layouts and instruction optimizations, and organized performance measurements and output-validation results
• Execution-time improvement: After instruction optimization, execution time decreased by 21.55% for brightness adjustment, 31.89% for binarization, and 27.94% for blur processing compared with the pre-optimization baseline
• Instruction analysis: Confirmed a 75% reduction in input load instructions in the image-processing section and compared changes in total instruction and branch instruction counts
• Output validation: Found zero byte mismatches and mean squared error (MSE) of 0 between the compared outputs for all three image-processing functions, confirming identical output after optimization

View presentation (PDF) ↗

Processor and Cache Performance Analysis Using SimpleScalar

Soongsil University · Computer Architecture course project (team)

Role: mcf benchmark analysis

• Performance analysis: Compared and visualized the mcf benchmark’s IPC and CPI across ALU counts, instruction issue widths, branch predictors, and cache configurations
• Processor configuration comparison: Increasing the instruction issue width (Scalar width) from 1 to 4 yielded a 39.8% increase in IPC
• Cache configuration analysis: Found that the relative performance of LRU and Random replacement policies changed with block size, and analyzed interactions among settings

View project report (PDF) ↗

Excellence Award, 2026 Project Report Challenge (CPC)

Soongsil University

ARM Instruction-Based Image Processing Optimization

Ignight Capital

Research Intern · Korea

• Protocol analysis: Reviewed technical documents and major updates for DeFi protocols; documented fund flows, collateral and staking structures, and version-specific changes in research reports
• Revenue and usage conditions: Compared revenue sources, incentives, liquidity, and fees across platforms; recorded transaction costs and processing times for deposits and withdrawals
• Prediction-market strategy research: Studied trading and reward structures and strategies for identifying arbitrage and other opportunities; assessed conditions and limits related to price volatility, liquidity, and data dependence
• Bot development support: Contributed to the development of arbitrage and prediction-market opportunity-scanning bots, using Claude as a coding aid.

View research samples (3 documents, PDF) ↗

Mobile Game Development — Planet10 & PlanetPop

Team Project

• Collaborative development: Co-developed two mobile puzzle games based on established game mechanics, contributing to feature planning, UI/UX design, and implementation
• Feature extensions: Contributed to ticket rewards and collectible planet cards in Planet10, and four game modes, gameplay items, and planet customization in PlanetPop
• Release outcomes: Released both games on Google Play and the App Store

View both games (App Store & Google Play) ↗