Soungmin (Min) Lee

Soungmin (Min) Lee

B.S./M.S. CS @ Georgia Tech

Publications

Conference Papers

Towards Truly Multilingual ASR: Generalizing Code-Switching ASR to Unseen Language Pairs

Published in ICML Workshop on Machine Learning for Audio, 2026

Investigates whether code-switching ASR capabilities learned from limited seen language pairs can generalize to unseen language pairs using model merging and domain generalization methods.

Recommended citation: Gio Paik, Hyunseo Shin, Soungmin Lee. (2026). "Towards Truly Multilingual ASR: Generalizing Code-Switching ASR to Unseen Language Pairs." ICML 2026 Workshop on Machine Learning for Audio. arXiv:2606.05846.
Download Paper

HiKE: Hierarchical Evaluation Framework for Korean-English Code-Switching Speech Recognition

Published in Findings of EACL, 2026

Introduces HiKE, the first globally accessible hierarchical benchmark for Korean-English code-switching speech recognition, featuring loanword-aware annotations and multi-level code-switching labels that enable fine-grained evaluation of multilingual ASR models.

Recommended citation: Gio Paik, Yongbeom Kim, Soungmin Lee, Sangmin Ahn, Chanwoo Kim. (2026). "HiKE: Hierarchical Evaluation Framework for Korean-English Code-Switching Speech Recognition." Findings of EACL 2026.
Download Paper | Download Slides

Words That Unite The World: A Unified Framework for Deciphering Central Bank Communications Globally

Published in NeurIPS, 2025

Introduces the World Central Banks (WCB) dataset, the largest multilingual corpus of monetary policy communications, with expert-annotated stance, temporal, and uncertainty labels. Benchmarks show cross-bank training improves classification and interpretability.

Recommended citation: Agam Shah, Siddhant Sukhani, Huzaifa Pardawala, Saketh Budideti†, Riya Bhadani†, Rudra Gopal†, Siddhartha Somani†, Michael Galarnyk†, Soungmin Lee†, et al. (2025). "Words That Unite The World: A Unified Framework for Deciphering Central Bank Communications Globally." NeurIPS 2025.
Download Paper | Download Slides