Khai Le-Duc @EMNLP 2025

MASc Student at U of Toronto | UHN | Ex RWTH i6

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

🚀 MultiMed-ST got into EMNLP 2025!

The world’s FIRST medical speech translation dataset - and the largest multilingual ST dataset ever

• 5 languages (VI/EN/FR/DE/ZH)
• 290k samples
• 20 directions

Pls⭐our repo: https://t.co/aDBpDSwQQT
Paper: https://t.co/FAvzlujJyY

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

@AudioAndSpeech 🔥 Big news! Our framework Sentiment Reasoning got into ACL Industry Track 2025 (Oral) 🎉
It lets AI decode hidden human emotions from speech + text 🧠💬💓

⭐ GitHub pls give us 1 star: https://t.co/KMRdnZyYFd
📄 Paper: https://t.co/gLZ9M1U5qm

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

🔥 Big news! Our framework Sentiment Reasoning got into ACL Industry Track 2025 (Oral) 🎉
It lets AI decode hidden human emotions from speech + text 🧠💬💓

⭐ GitHub pls give us 1 star: https://t.co/G9o0njnAbz
📄 Paper: https://t.co/biP9rHNYtb

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

💫We would like to announce that our work "MultiMed: Multilingual Medical Speech Recognition via Attention Encoder Decoder" has been accepted to ACL 2025.

🌐 Link to paper: https://t.co/Qut39azLPo

🌐GitHub: https://t.co/hx6DMWzdkR

Pls press ⭐️ on our repo if you feel helpful

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

💟 We would like to announce our paper "Medical Spoken Named Entity Recognition" to be presented at NAACL 2025

✨ Paper: https://t.co/pZz8ic2flK
✨ Code & dataset: https://t.co/Q1LuYsncKh

💫 Please press star ⭐ if you find my Github repo useful 😊

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

Information retrieval from speech?
Enjoy our new preprint/demo here: https://t.co/7TkGAqA9nL

[arXiv Sound @ArxivSound]

``wav2graph: A Framework for Supervised Learning Knowledge Graph from Speech,'' Khai Le-Duc, Quy-Anh Dang, Tan-Hanh Pham, Truong-Son Hy, https://t.co/e1Zv8BQuIL

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

My paper has been accepted at Interspeech 2024.
We propose the first real-time speech summarization system deployable in industry

[arXiv Sound @ArxivSound]

``Real-time Speech Summarization for Medical Conversations,'' Khai Le-Duc, Khai-Nguyen Nguyen, Long Vo-Dang, Truong-Son Hy, https://t.co/gMlKXb3Mjm

[arXiv Sound @ArxivSound]

``Medical Spoken Named Entity Recognition,'' Khai Le-Duc, https://t.co/Z2PuQG1YPk

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

V2 of the VietMed paper has been released!

[Audio and Speech Processing Papers @AudioAndSpeech]

VietMed: A Dataset and Benchmark for Automatic Speech Recognition of Vietnamese in the Medical Domain. https://t.co/LdFWoi0k6D

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

My VietMed paper was just presented at LREC-COLING 2024.
VietMed is the world's largest medical ASR dataset in all 7 benchmarks.
For the first time in history, the world's largest dataset is named Vietnam.
https://t.co/he1fDu7al8
https://t.co/xt5yS2iSrT
https://t.co/j3o7Otb18E

[Khai Le-Duc @EMNLP 2025 @_leduckhai_]

Check out my very first first-author paper at LREC-COLING 2024 😁
✨️VietMed is by far the world's largest medical speech recognition dataset in 7 aspects

[Audio and Speech Processing Papers @AudioAndSpeech]

VietMed: A Dataset and Benchmark for Automatic Speech Recognition of Vietnamese in the Medical Domain. https://t.co/LdFWoi0k6D