Jun Woo Chung

Implemented novel methods for approximate nearest neighbor search graph index building. Developed and implemented new algorithms improving graph index maintenance efficacy of state-of-the-art. Designed novel algorithms for attribute-aware filtered nearest neighbor search to enhance query precision and speed.

Graduate Research Assistant at Rochester Institute of Technology (2022-08 – Present)

• Implemented novel methods for approximate nearest neighbor search graph index building.
• Developed and implemented new algorithms improving graph index maintenance efficacy of state-of-the-art.
• Designed novel algorithms for attribute-aware filtered nearest neighbor search to enhance query precision and speed.
• Implemented robust watermarking techniques into gradient boosting models for intellectual property protection.
• Devised data-free poison detection protocols for GBDTs, securing models against adversarial training inputs.

Technical Intern at Waymo (2026-05 – 2026-08)

• Designed and formulated a multivariate distributional realism metric using multi-dimensional Optimal Transport to evaluate the fidelity of simulated autonomous agent behaviors against large-scale real-world driving logs.
• Engineered high-performance computation pipelines and benchmarking suites to validate metric sensitivity across baseline configurations, establishing a unified topline scorecard to guide simulation model hillclimbing and release validation.

Student Assistant at Lawrence Berkeley National Laboratory (2024-05 – 2025-03)

• Implemented machine learning models to analyze and predict REE concentrations from multisensor data of mine tailings.
• Developed methods to improve model accuracy with limited sample data.
• Evaluated the effects of missing modalities and spatial features on model generalizability.
• Investigated physical inconsistencies and inter-site differences in REE measurement data, identifying critical challenges in generalization and measurement consistency.

Student Assistant at Lawrence Berkeley National Laboratory (2023-05 – 2023-08)

• Conducted case study on spectral data preparation for machine learning, specifically focusing on gamma-ray spectral data.
• Implemented data pre-processing strategies and enhancement of machine learning models for classifying geological types from gamma-ray spectral data, increasing classification accuracy from 40.8% to 72.7%.

Researcher (alternative military service) at Silicon Display, co. Ltd. (2018-10 – 2022-03)

• Developed an automated database management and quantitative analysis system for TFT fingerprint sensor production, enhancing efficiency and accuracy.
• Performed circuit design validation and optimization in over 10 sensor development projects.
• Created over 50 scripts for CAD design automation and fingerprint sensor image analysis, streamlining workflow and data processing.
• Engineered an automatic fitting program for experimental data in TFT simulation modeling.
• Conducted statistical analysis of acceptance tests across 3 production lines, identifying key process steps and design flaws.
• Applied image and statistical analysis to detect production issues in 20+ fingerprint sensor projects, enhancing product reliability.
• Optimized TFT device performance through geometry and deposition process refinement using finite element analysis.
• Led a 6-month collaborative R&D project on low-temperature polycrystalline silicon TFTs, achieving process innovations and efficiency improvements.

Summer Undergraduate Research Fellowship at California Institute of Technology (2016-06 – 2016-08)

• Analyzed physical characteristics of nanobeam optomechanical cavities via finite element method using COMSOL Multiphysics, Python and Matlab.
• Presented results in seminar of 30+ undergraduate students and supervisors.

Ph.D. in Computer Science – Rochester Institute of Technology (2022-08 – 2027-05)

M.Sc. in Physics – University of Cambridge (2016-10 – 2017-06)

B.A. in Physics – University of Cambridge (2013-10 – 2016-06)