Tzu-Ming Harry Hsu

Founder and CTO, Hashgreen Labs
Ph.D in Computer Science, Massachusetts Institute of Technology

Professional Skills

Business Product Design General Management Company Growth

Development Management

Research Interests

Computer Vision Federated Learning ML for Healthcare

Machine Learning Signal Processing


  • 2 advanced degrees: Ph.D. and Master's obtained from Massachusetts Institute of Technology
  • 2 majors acquired in undergraduate education: Electrical Engineering and Physics
  • Ranked #1 in International Physics Olympiad 2011 in both theory and experiment among 400+ participants


Hashgreen Labs.
Founder & CTO

Sep 2022 - Now


Massachusetts Institute of Technology
Ph.D. Student in Electrical Engineering and Computer Science

Sep. 2017 - Aug. 2022

Massachusetts Institute of Technology
S.M. in Electrical Engineering and Computer Science

Sep. 2017 - May. 2020

National Taiwan University (NTU) Class Rank: 1/190 GPA: 3.99/4.00

Sep. 2011 - Jun. 2016


Emulating Clinical Diagnostic Reasoning for Jaw Cysts with Machine Learning

Balazs Feher, Ulrike Kuchler, Falk Schwendicke, Lisa Schneider, Jose Eduardo Cejudo Grano de Oro, Tong Xi, Shankeeth Vinayahalingam, Tzu-Ming Harry Hsu, Janet Brinz, Akhilanand Chaurasia, Kunaal Dhingra, Robert Andre Gaudin, Hossein Mohammad-Rahimi, Nielsen Pereira, Francesc Perez-Pastor, Olga Tryfonos, Sergio E Uribe, Marcel Hanisch, Joachim Krois

Artificial Intelligence to Assess Body Composition on Routine Abdominal CT Scans and Predict Mortality in Pancreatic Cancer – A Recipe for Your Local Application

Tzu-Ming Harry Hsu, Khoschy Schawkat, Seth J. Berkowitz, Jesse L. Wei, Alina Makoyeva, Kaila Legare, Corinne DeCicco, S. Nicolas Paez, Jim S.H. Wu, Peter Szolovits, Ron Kikinis, Arthur J. Moser, Alexander Goehler
European Journal of Radiology

Visceral Adiposity and Severe COVID-19 Disease: Application of an Artificial Intelligence Algorithm to Improve Clinical Risk Prediction

Alexander Goehler, Tzu-Ming Harry Hsu, Jacqueline A. Seiglie, Mark J. Siedner, Janet Lo, Virginia Triant, John Hsu, Andrea Foulkes, Ingrid Bassett, Ramin Khorasani, Deborah J. Wexler, Peter Szolovits, James B. Meigs, Jennifer Manne-Goehler
Open Forum Infectious Diseases

Adversarial Contrastive Pre-training for Protein Sequences


Matthew McDermott, Brendan Yap, Tzu-Ming Harry Hsu, Di Jin, Peter Szolovits

Three-Dimensional Neural Network to Automatically Assess Liver Tumor Burden Change on Consecutive Liver MRIs


Alexander Goehler, Tzu-Ming Harry Hsu, Ronilda Lacson, Isha Gujrathi, Raein Hashemi, Grzegorz Chlebus, Peter Szolovits, Ramin Khorasani
Journal of the American College of Radiology

Automatic Longitudinal Assessment of Tumor Responses


Tzu-Ming Harry Hsu
Master's Thesis, MIT EECS

Baselines for Chest X-Ray Report Generation


William Boag, Tzu-Ming Harry Hsu, Matthew McDermott, Gabriela Berner, Emily Alesentzer, Peter Szolovits
Machine Learning for Health (ML4H) Workshop, NeurIPS 2019
* Poster by William Boag

Clinically Accurate Chest X-Ray Report Generation

arXiv Code Poster

Tzu-Ming Harry Hsu*, Guanxiong Liu*, Matthew McDermott, Willie Boag, Wei-Hung Weng, Peter Szolovits, Marzyeh Ghassemi
Machine Learning for Healthcare 2019
* Co-first Authors; Spotlight Presentation

Transfer Neural Trees: Semi-Supervised Heterogeneous Domain Adaptation and Beyond

Wei-Yu Chen, Tzu-Ming Harry Hsu, Yao-Hung Hubert Tsai, Ming-Syan Chen, and Yu-Chiang Frank Wang
IEEE Transactions on Image Processing (TIP)

3D-Aware Scene Manipulation via Inverse Graphics

arXiv Project Code Poster

Tzu-Ming Harry Hsu*, Shunyu Yao*, Jun-Yan Zhu, Jiajun Wu, Antonio Torralba, William T. Freeman, Joshua B. Tenenbaum
NeurIPS 2018

Learning Food Quality and Safety using Wireless Stickers

Unsoo Ha, Yunfei Ma, Zexuan Zhong, Tzu-Ming Harry Hsu, Fadel Adib
Hotnets 2018

Transfer Neural Trees for Heterogeneous Domain Adaptation

Wei-Yu Chen, Tzu-Ming Harry Hsu, Yao-Hung Hubert Tsai, and Yu-Chiang Frank Wang
ECCV 2016

Unsupervised Domain Adaptation With Imbalanced Cross-Domain Data

Tzu-Ming Harry Hsu, Wei-Yu Chen, Cheng-An Hou, Yao-Hung Hubert Tsai, Yi-Ren Yeh, and Yu-Chiang Frank Wang
ICCV 2015

Connecting the dots without clues: Unsupervised domain adaptation for cross-domain visual classification

Wei-Yu Chen, Tzu-Ming Harry Hsu, Cheng-An Hou, Yi-Ren Yeh and Yu-Chiang Frank Wang
ICIP 2015

Robust Motion Artifact Reduction of Photoplethysmographic Signal with Trajectory Space Circular Model

Tzu-Ming Harry Hsu, Wei-Yu Chen, Kuan-Lin Chen, Mong-Chi Ko, You-Cheng Liu, An-Yeu Andy Wu
ICASSP Signal Processing Cup 2015

Research Experiences

Clinical Descision Making Group, CSAIL, MIT
Prof. Peter Szolovits

Jul. 2018 - Now

Mobile Vision Group, Google Research
Matthew Brown

Jun. 2019 - Mar. 2020

Computer Vision Labs, CSAIL, MIT

Jan. 2018 - Jul. 2018
  • Vision as Inverse Graphics

    Understand scenes with approximately differentiable inverse graphics + graphics renderer network.

Signal Kinetics Lab, Media Lab, MIT
Prof. Fadel Adib

Sep. 2017 - Jan. 2018
  • Agent Co-localization via Cellular Network

    Construct Bayesian model based on LTE channel estimation and locate agents in an area simultaneously.

Multimedia and Machine Learning Lab, Academia Sinica
Prof. Yu-Chiang Frank Wang

Apr. 2014 - Jun. 2016
  • Deep Learning for Heterogeneous Domain Adaptation

    Transfer knowledge across different feature domains and build classifiers above the transferred knowledges.

    Transfer Neural Trees is proposed to transfer classifiers to a different dimensional space with deep neural network.

  • Unsupervised Domain Adaptation with Imbalanced Cross-domain Data

    Information of labeled source-domain data is transferred to the unlabeled target-domain, which may be a small set with imbalanced label counts.

    Closest Common Space Learning is proposed to combine sub-domain level classifiers to identify better source data applicability.

  • Unsupervised Domain Adaptation with Balanced Cross-domain Data

    A set of labeled source-domain data is used to construct classifier for the unlabeled target-domain data.

    An algorithm is proposed to address source-target mismatch and project them to a common space.

  • External Review

    Review papers as external reviewer for ICCV, ECCV, AAAI, and IJCAI.

Sep. 2014 - Jun. 2015
  • Noise Removal of Photoplethysmographic (PPG) Signals

    Remove noises in PPG signals induced by motions by decorrelating the PPG with accelerometer signal.

    An algorithm is proposed to project the signal into a complex plane, in which a temporal filter will be performed, followed by ensemble voting for the optimal beat counts.

Laboratory for Applied Logic and Computation in System Design (ALCom Lab)
Prof. Jie-Hong Roland Jiang

Jul. 2013 - Jun. 2014
  • Compressed Sensing

    Compress the data perceived by a sensor array using less data storage than what it used to consume.

  • Mathematical Neural Models

    Establish a time-continuous model of human neurons to simulate the biological effects at stimulus and message passing.

Honors & Awards


Altera Innovate Asia FPGA Design Competition
Silver Medal Award


Ranked 2nd among 20 teams, team Taipei Amoeba designed a custom PCB named EZBud with algorithms integrated inside, which communicates with the FPGA. This piece of hardware modulates music according to measured user sporting statistics.

ICASSP Signal Processing Cup
Tenth Place


Ranked 10th globally in sports heartbeat detection with an error of 4.89 beats per minute (BPM), team Taipei Amoeba had proposed an algorithm called Trajectory Space Circular Model.


Presidential Award (5 times)
Issued by the Department of Electrical Engineering

2011 - 2014

Awarded per semester to the top 5% students.

International Physics Olympiad (IPhO)
World’s First Place and Gold Medal Award


Ranked 1st in both theory section and experiment section among 401 national representatives from senior high schools of over 80 countries.

International Junior Science Olympiad (IJSO)
Gold Medal Award


Ranked top 10% among 300 national representatives from junior high schools of over 60 countries.

Work & Teaching Experiences

Research Intern

Jun. 2019 - Aug. 2019

Digital Drift Corporation

Mar. 2016 - Now
Deep Neural Networks for Recognition and Matching

Build deep models for cuisine images using TensorFlow on multi-GPU machines, providing a backend with an API.

Olympiad Tutoring Community

Sep. 2011 - Jun. 2015

Offer tutoring for high school physics, competition physics, GRE subject test (physics), and SAT II subject test (physics).

Two students became national representatives for Taiwan in International Physics Olympiad (IPhO).