Hao CHEN (@haochen_hkust) 's Twitter Profile
Hao CHEN

@haochen_hkust

Assistant Professor in HKUST CSE

ID: 1392479814716510209

linkhttps://cse.hkust.edu.hk/~jhc/ calendar_today12-05-2021 14:00:41

87 Tweet

200 Followers

305 Following

HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

๐Ÿš€ HKUST Smart Lab showcased our groundbreaking AI foundation models for healthcare, GSCo and mSTAR, to NVIDIA AI's Jensen Huang! Better things are on the wayโ€”learn more about these innovations above! #Healthcare #HKUST mSTAR: hkustsmartlab.github.io/2024/07/31/mstโ€ฆ GSCo: arxiv.org/abs/2404.15127

๐Ÿš€ HKUST Smart Lab showcased our groundbreaking AI foundation models for healthcare, GSCo and mSTAR, to <a href="/NVIDIAAI/">NVIDIA AI</a>'s Jensen Huang! Better things are on the wayโ€”learn more about these innovations above! #Healthcare #HKUST
mSTAR: hkustsmartlab.github.io/2024/07/31/mstโ€ฆ
GSCo: arxiv.org/abs/2404.15127
HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

New study: Hierarchical Multi-Instance Learning (HMIL) for fine-grained WSI classification accepted by IEEE-TMI! ๐ŸŽ‰ ๐Ÿ” HMIL improves cancer diagnosis by leveraging hierarchical label correlations, outperforming SOTA methods on multiple datasets. More: hkustsmartlab.github.io/2024/12/27/hmiโ€ฆ

New study: Hierarchical Multi-Instance Learning (HMIL) for fine-grained WSI classification accepted by IEEE-TMI! ๐ŸŽ‰

๐Ÿ” HMIL improves cancer diagnosis by leveraging hierarchical label correlations, outperforming SOTA methods on multiple datasets.

More: hkustsmartlab.github.io/2024/12/27/hmiโ€ฆ
Hao CHEN (@haochen_hkust) 's Twitter Profile Photo

Check out our new IEEE TMI paper HMIL. HMIL improves cancer diagnosis by leveraging hierarchical label correlations, outperforming SOTA methods on multiple datasets.

HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

Smart Lab unveils Smart-CCS! An AI system using pretraining & test-time adaptation for accurate cervical cancer screening. Validated on 20 retrospective & prospective datasets, it ensures high sensitivity & generalization in real-world settings. Read more: hkustsmartlab.github.io/2025/02/17/smaโ€ฆ

Smart Lab unveils Smart-CCS! An AI system using pretraining &amp; test-time adaptation for accurate cervical cancer screening. Validated on 20 retrospective &amp; prospective datasets, it ensures high sensitivity &amp; generalization in real-world settings.
Read more: hkustsmartlab.github.io/2025/02/17/smaโ€ฆ
HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

Our paper FOCUS is accepted at #CVPR2025! ๐ŸŽ‰ FOCUS introduces knowledge-enhanced adaptive visual compression for few-shot WSI classification, leveraging foundation models & language priors to improve diagnostic accuracy. Read more: hkustsmartlab.github.io/2025/02/28/focโ€ฆ #AI #Pathology #HKUST

Our paper FOCUS is accepted at #CVPR2025! ๐ŸŽ‰ FOCUS introduces knowledge-enhanced adaptive visual compression for few-shot WSI classification, leveraging foundation models &amp; language priors to improve diagnostic accuracy. 
Read more: hkustsmartlab.github.io/2025/02/28/focโ€ฆ
#AI #Pathology #HKUST
Hao CHEN (@haochen_hkust) 's Twitter Profile Photo

Please check out our 2025 CVPR paper FOCUS on knowledge-enhanced adaptive visual compression for few-shot WSI classification.

Hao CHEN (@haochen_hkust) 's Twitter Profile Photo

Check out Smart-CCS! An AI system using pretraining & test-time adaptation for accurate cervical cancer screening. Validated on 20 retrospective & prospective datasets, it ensures high sensitivity & generalization in real-world settings.

IEEE-TMI (@ieee_tmi) 's Twitter Profile Photo

๐Ÿš€New paper alert! โžก๏ธLLM ๐˜„๐—ถ๐˜๐—ต ๐—ฅ๐—ฒ๐—ด๐—ถ๐—ผ๐—ป-๐—ด๐˜‚๐—ถ๐—ฑ๐—ฒ๐—ฑ ๐—ฅ๐—ฒ๐—ณ๐—ฒ๐—ฟ๐—ฟ๐—ถ๐—ป๐—ด ๐—ฎ๐—ป๐—ฑ ๐—š๐—ฟ๐—ผ๐˜‚๐—ป๐—ฑ๐—ถ๐—ป๐—ด ๐—ณ๐—ผ๐—ฟ ๐—–๐—ง ๐—ฅ๐—ฒ๐—ฝ๐—ผ๐—ฟ๐˜ ๐—š๐—ฒ๐—ป๐—ฒ๐—ฟ๐—ฎ๐˜๐—ถ๐—ผ๐—ป The Reg2RG framework leverages segmentation masks to focus on diagnostic regions! Read the paper: ๐Ÿ”—ieeexplore.ieee.org/document/10963โ€ฆ

๐Ÿš€New paper alert!

โžก๏ธLLM ๐˜„๐—ถ๐˜๐—ต ๐—ฅ๐—ฒ๐—ด๐—ถ๐—ผ๐—ป-๐—ด๐˜‚๐—ถ๐—ฑ๐—ฒ๐—ฑ ๐—ฅ๐—ฒ๐—ณ๐—ฒ๐—ฟ๐—ฟ๐—ถ๐—ป๐—ด ๐—ฎ๐—ป๐—ฑ ๐—š๐—ฟ๐—ผ๐˜‚๐—ป๐—ฑ๐—ถ๐—ป๐—ด ๐—ณ๐—ผ๐—ฟ ๐—–๐—ง ๐—ฅ๐—ฒ๐—ฝ๐—ผ๐—ฟ๐˜ ๐—š๐—ฒ๐—ป๐—ฒ๐—ฟ๐—ฎ๐˜๐—ถ๐—ผ๐—ป

The Reg2RG framework leverages segmentation masks to focus on diagnostic regions!

Read the paper: ๐Ÿ”—ieeexplore.ieee.org/document/10963โ€ฆ
HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

๐ŸŽ‰ New in IEEE-TMI: Reg2RG from HKUST Smart Lab! A region-guided large language model that generates accurate, interpretable chest CT reports by linking diagnostic text to specific anatomical regions. More: smartlab.cse.ust.hk/2025/05/21/regโ€ฆ #MedAI #Radiology #LLM

๐ŸŽ‰ New in <a href="/IEEE_TMI/">IEEE-TMI</a>: Reg2RG from HKUST Smart Lab!
A region-guided large language model that generates accurate, interpretable chest CT reports by linking diagnostic text to specific anatomical regions.
More: smartlab.cse.ust.hk/2025/05/21/regโ€ฆ
#MedAI #Radiology #LLM
HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

Excited our #mSTAR (Lead by YingxueXu & Yihui Wang) & #SmartPath AI for pathology are featured in nature! At #HKUST, we focus on real clinical impact & explainable #AIinMedicine. The journeyโ€™s just starting! Read: doi.org/10.1038/d41586โ€ฆ #DigitalPathology

HKUST Smart Lab (@smartlab_hkust) 's Twitter Profile Photo

Meet SmartCare! ๐Ÿš€ Developed by SmartLab (Lead by S. He and Y. Nie Yuxiang Nie) at HKUST. Now pilot-testing at the campus clinic with 15,000+ students & staff. Smarter triage, easier notes, better care. #AI #Healthcare Read more: smartlab.cse.ust.hk/2025/07/29/smaโ€ฆ

Meet SmartCare! ๐Ÿš€ Developed by SmartLab (Lead by S. He and Y. Nie <a href="/npyuxl/">Yuxiang Nie</a>) at HKUST. Now pilot-testing at the campus clinic with 15,000+ students &amp; staff. Smarter triage, easier notes, better care. #AI #Healthcare

Read more: smartlab.cse.ust.hk/2025/07/29/smaโ€ฆ