Explainable AI-powered Baseball Swing Coaching System
오승현, 황준수 / Seunghyun Oh, Junsoo Hwang
Proceedings of HCI Korea 2026, pp. 1518–1523, Feb. 2026.
Equal contribution.
This paper proposes STAR-Coach, an explainable AI-powered baseball swing coaching system. STAR-Coach reinterprets refined pose outputs as a swing template and tracks correction actions to provide actionable feedback on which body parts should be corrected, when correction is needed, how much adjustment is required, and how feedback can be presented visually, quantitatively, and textually.
Research connection: This work connects my current direction in XAI and human-verifiable AI systems with practical coaching feedback. It shows my interest in moving beyond prediction accuracy toward explanations that users can understand and act on.
Search-Pro: Schema-Grounded Text-to-SQL Assistant
Search-Pro is a sanitized public snapshot of a Text-to-SQL assistant. It translates natural-language questions into a schema-grounded semantic plan, renders read-only SQL on the server side, validates query shape, executes against a mock database pipeline, and records traces for human verification.
Research angle: Text-to-SQL is not only a generation problem. It is also an explainability and evaluation problem: users need to understand why a query was generated, whether it is safe, and how to detect wrong but plausible results.
Backend Engineer → XAI / Agent Explainability Researcher-in-Progress
Turning practical AI engineering experience into testable research questions.
저는 완성된 연구자가 아니라, 백엔드 개발자로 쌓은 실무 경험을 작고 검증 가능한 연구 질문으로 바꾸는 과정에 있습니다. ICML 2026에서는 논문 저자 포지션보다, XAI, medical imaging interpretability, uncertainty, agent explainability 연구자들과 대화하며 방향을 좁히는 것이 목표입니다.