Academic & professional background

Curriculum vitae

Research in speech recognition and audio-language models, alongside experience in applied machine learning and embedded systems.

Education

Sogang University, Seoul, Korea

Mar. 2026 – Present

M.S. Student, Department of Artificial Intelligence

Intelligent Information Processing Lab (IIP) · Advisor: Prof. Hyung-Min Park

Inha University, Incheon, Korea

Feb. 2019 – Aug. 2025

B.S. in Electronic Engineering

GPA 3.98 / 4.5

Research experience

Graduate Researcher

Mar. 2026 – Present

Intelligent Information Processing Lab (IIP), Sogang University

  • Study lightweight and robust ASR with Zipformer and Conformer, including medical-domain speech recognition (SCH), quantization, and knowledge distillation.
  • Investigate ASR integration with language models and spatial and acoustic understanding from multichannel microphone inputs.
  • Co-authored a DCASE 2026 Task 5 technical report on task-leaf routing with MiMo-Audio for audio-dependent question answering.

Research Intern

Jul. 2025 – Feb. 2026

Intelligent Information Processing Lab (IIP), Sogang University · Presentation (PDF)

  • Adapted Voxtral Mini 3B for Korean medical ASR (SCH) and spoken question answering.
  • Curated medical text/audio datasets and applied two-stage fine-tuning with 4-bit QLoRA.
  • Project evaluation against the Voxtral 3B 4-bit baseline: CER 19.14% → 9.09%; GPT-OSS-judged spoken QA 2.07 → 4.01/5.

Research Intern

Jan. 2024 – Dec. 2024

Machine Intelligence Lab, Inha University

  • Investigated scene-graph-conditioned video generation with diffusion models.
  • Implemented an autoregressive module for long video generation.

Research Intern

Jul. 2022 – Dec. 2022

Intelligence Embedding System Lab (IESL), Inha University

  • Conducted computer vision research for robotics and autonomous driving.
  • Co-authored a KSAE 2022 poster on ROS-based environmental perception and decision-making for indoor autonomous mobile robots.

Publications & reports

  1. Jongha Kim, Leehyeon Song, Hyung-Min Park. “Task-Leaf Routed MiMo-Audio for DCASE 2026 Task 5.” DCASE 2026 Challenge, technical report, Jun. 2026. Official evaluation accuracy: 49.13% (Kim_SGU_task5_4). Report (PDF) · Challenge results

  2. Jang Ji-hye, Lee Young-jun, Heo Ji-won, Kim Jong-ha. “Development of an ROS-based Environmental Perception and Decision-making System for Indoor Autonomous Mobile Robots.” Korean Society of Automotive Engineers (KSAE), Jeju, Korea, Oct. 2022. (Poster)

Work experience

ML Scientist / Embedded Engineer

Mar. 2025 – Jun. 2025

R&D Department, STS Engineering

  • Developed soft-sensor models to predict water quality indicators from IoT sensor time series.
  • Built MCU firmware on FreeRTOS for continuous unattended operation, with fault handling and reboot recovery.
  • Delivered a monitoring prototype; reported project NMAE decreased from 15% to 8%.

Selected projects

All projects ↗

Audio-to-Audio Speech Unit Translation (Korean ↔ English)

Sep. 2024 – Present

ECE Capstone Design, Inha University · GitHub

Exploring Korean–English speech-to-speech translation through discrete speech units, without an intermediate text representation at inference time.

Generated English speech scored BLEU 42.8 / COMET 0.1009 in the project evaluation. Output was intelligible but had unstable pitch and background noise.

Scene Graph to Video Generation with Diffusion

Aug. 2024 – Dec. 2024

Machine Intelligence Lab, Inha University

Explored controllable video generation with scene graphs as structured conditions for a diffusion model.

Implemented the full pipeline. Generation quality remained limited by scarce scene-graph–video paired data (Action Genome) and weak graph–video alignment.

Competitions

Korean LLM Fine-tuning for Question Answering

Jul. 2024 – Aug. 2024

2024 Inha Artificial Intelligence Challenge (Dacon) · GitHub

Explored question answering over Korean economic articles through language-model fine-tuning.

LoRA adaptation degraded the already well-aligned base model in project experiments; the approach shifted to minimal fine-tuning to preserve base behavior.

Global Wildfire Detection Challenge

Mar. 2024

6th AI SPARK Challenge · GitHub

Built a satellite-image segmentation system for wildfire detection using an ensemble of TransUNet and Attention U-Net.

Teaching

Teaching Assistant

Sep. 2026 – Present

Sogang University, Dept. of Electronic Engineering

Semiconductor Elite Project (EEE3150)

  • Designed the course project: a lightweight streaming ASR system with end-point detection and LLM integration.

Teaching Assistant

Mar. 2026 – Jun. 2026

Sogang University, Dept. of Electronic Engineering

Advanced Engineering Mathematics I (EEE2103)

Teaching Assistant

Sep. 2024 – Dec. 2024

Inha University, Dept. of Electronic Engineering

Deep Learning · Introduction to Machine Learning

Entrepreneurship

  • Startup: Product Development and Branding — Gyeonggi Content Agency, 20M KRW funding (Jan. – Jul. 2023).
  • SeTA (Social Entrepreneurship Team Academy) — SKKU / MTA Korea, global entrepreneurship training program (Mar. – Jun. 2022).