CV

click here for pdf file

Professional Profile

Lead 2 projects as principal investigator, targeting the development of transformative technology for medical and healthcare service
Raised $1M (USD) in funding with 2 research proposals within 2 years
Manage 3 research projects as an AI engineer lead for the development of deep learning / machine learning models for medical and healthcare services
Selected to be one of 30 alumni coaches as part of the inaugural IDEO U Alumni Fellowship program to engage and guide IDEO U online learners

Education

M.S. in Occupational Therapy, National Taiwan University, 2013-2015
- Supervisor: Prof. Hao-Ling Chen
- Research Focus: Computer Methods in Motion Analysis
- Master’s Thesis: The Effects of External Task Difficulties on Arm-Trunk Movements in Children with Hemiplegic Cerebral Palsy
B.S. in Occupational Therapy, National Taiwan University, 2009-2013
- Dean’s List * 2

Work & Research Experience

Artificial Intelligence Engineer, 06-2019-Present
- Taiwan AI Labs, Taipei, Taiwan
- Projects:
  1. Pretrained Backbones for Medical Images
    - Developed CNN-based pretrained backbones for all kinds of CT and MRI images.
    - This aims to boost up the overall performance for semantic segmentation on at least 10 datasets, including Brain, Heart, Liver, Hippocampus, Prostate, Lung, Hepatic Vessel, Pancreas, Spleen, Colon.
  2. General Medical Image Dataset Framework for Deep Learning Modeling Pipeline
    - Designed a general module to deal with all kinds of CT and MRI images.
    - This enabled AI engineers to process medical images efficiently, which saved nearly 80% of the time.
Project Investigator & Artificial Intelligence Researcher, 04/2017-06/2019
- Service Systems Technology Center (SSTC), Industrial Technology Research Institute (ITRI), Hsinchu, Taiwan
- Projects:
  1. AI-based Early-stage Glaucoma Detection with Fundus Images
    - Period: 01/2018-12/2019
    - Developed an encoder-decoder CNN to predict the retinal nerve fiber layer defects based on fundus images. With GAN, the dice coefficient and correlation coefficient of the model increased (2%) compared to the baseline model (only encoder-decoder CNN)
    - 2 conference paper (1 submitted, 1 accepted)
  2. The Development of Medical Devices for COPD Exacerbation Development Monitoring
    - Period: 04/2017-12/2018
    - Designed an innovative service system to help patients with chronic obstructive pulmonary disease (COPD) detect acute exacerbation earlier by means of abnormal lung sound detection
    - Achieved 85.78% of accuracy for the model for lung sound detection, 5% higher than the accuracy reported in reference paper
    - filed 2 patents
Artificial Intelligence Engineer, 09/2015 - Present
- Service Systems Technology Center (SSTC), Industrial Technology Research Institute (ITRI), Hsinchu, Taiwan
- Projects:
  1. Eye Alpha Zero – The Development of Incremental Learning for AI in Ophthalmology
    - Period: 01/2019 - 12/2020
    - Aim to develop an incremental learning algorithm for AI in Ophthalmology
  2. Artificial Intelligence-based Diabetics Fundus Image Decision Support
    - Designed an innovative system to prevent patients with diabetes from severe diabetic retinopathy
    - Built model for classifying the severity of diabetic retinopathy with kappa score up to 84.15%, higher than that reported by Google and that of the 3rd place of Diabetic Retinopathy Detection Challenge in Kaggle
    - Collaborated with more than 3 hospitals to collect fundus images and annotate each image with possible DR lesions (microaneurysm, exudate, hemorrhage, neovascularization, etc). The results will be published soon.
IDEO U Alumni Coach Fellow (Volunteer), 01/2017-07/2017, 08/2018-Present
- Selected to be one of 30 Alumni Coaches as part of the inaugural IDEO U Alumni Fellowship program. Responsibilities include providing guidance to IDEO U learners, participating in IDEO U workshops and building a global community of practice.
- IDEO U Alumni Fellows met the following requirements for selection: provided high quality, enriching contributions to the IDEO U community as learners, demonstrated a robust understanding of the content, and have worked to apply the learnings in their own contexts.
Research Assistant, 09/2013-08/2015
- Movement Science and Assistive Technology Lab, School of Occupational Therapy, College of Medicine, National Taiwan University
- Projects:
  1. Development and Application of Rehabilitation Games for Children with Cerebral Palsy Using Motion Sensing Technology
    - Designed a Kinect-based rehabilitation program using modern therapeutic approach
    - Collaborated with cross-disciplinary team, including occupational therapist, industrial designer and software engineer.
Teaching Assistant, 09/2013-06/2015
- School of Occupational Therapy, College of Medicine, National Taiwan University
- Help graduate students learn programming in Matlab
- Coach international occupational therapist intern to do physical assessments on stroke patients
- Teach undergraduates to identify human anatomy from Cadaver
Occupational Therapy Intern, 07/2012-06/2013
- National Taiwan University Hospital (NTUH)
- Full-time internship in NTUH. Learn to be an independent occupational therapist in following 3 fields: pediatric, physical and psychological.
- Got certificate for being an occupational therapist in 09/2013

Skills

Python
- use Pandas + Matplotlib for data preprocessing, statistical analysis and visulaization
- use Flask with uWSGI to build RESTful API for the interface of AI inference
- use OpenCV and PIL for image preprocessing
- use LibROSA for audio preprocessing, STFT, MFCC and spectrogram generation
Deep Learning / Machine Learning
- Skill at modeling with Pytorch and Tensorflow
- Scikit-learn for machine learning models
C++ / HTML / CSS / SQL
Design Thinking / User-Centered Design

Publications

Postural control during standing reach in children with Down syndrome

Chen, H.-L., Yeh, C.-F., & Howe, T.-H. (2015). Postural control during standing reach in children with Down syndrome. Res Dev Disabil.. 38(0), 345-351. doi: http://dx.doi.org/10.1016/j.ridd.2014.12.024

Body Motion Analysis System, Portable Device and Body Motion Analysis Method.

Chun-Fu Yeh, Szu-Han Tzao and Ming-Chieh Tsai. (2016) Body Motion Analysis System, Portable Device and Body Motion Analysis Method. U.S. Patent 2018/0178060A1, filed December 23, 2016. Patent Pending.

Feasibility of Early Detection for Retinal Nerve Fiber Layer Defect with Digital Fundus Image in Glaucoma Patients

Chun-Fu Yeh, Guan-An Chen, Kwou-Yeung Wu, Min-Yu Huang (2018) Feasibility of Early Detection for Retinal Nerve Fiber Layer Defect with Digital Fundus Image in Glaucoma Patients. The 31st IPPR Conference on Computer Vision, Graphics and Image Processing, Tainan, Taiwan, August 19-21.

LOCALIZATION OF RETINAL NERVE FIBER LAYER DEFECT IN FUNDUS IMAGE BY VISUAL FIELD GUIDED LEARNING NETWORK

Chun-Fu Yeh, Guan-An Chen, Kwou-Yeung Wu, Min-Yu Huang (2019) Localization of Retinal Nerve Fiber Layer Defect in Fundus Image by Visual Field Guided Learning Network. The 32st IPPR Conference on Computer Vision, Graphics and Image Processing, Taitung, Taiwan, August 25-27.

More Publications are listed in CV pdf file

Service and leadership

Lead two research projects (total funding: ~1M USD)
Director General of United Eagle Welfare Committee, 01/2017-12/2017
- Industrial Technology Research Institute (ITRI), Hsinchu, Taiwan
- Took charge of 300K USD budget to hold activities that can benefit employees

Training & Certificate

Computer Vision Nanodegree Program, 01/2019 - 04/2019
- Online Course, Udacity
- Learned to build computer vision applications based with Pytorch, completing following projects:
  1. Image Captioning with CNN plus RNN
  2. Landmark Detection & Tracking using SLAM
Immersive Silicon Valley Training (ISVT), 07/2018 - 09/2018
- San Francisco Bay Area, California, USA
- Designated by the director of Service Systems Technology Center, Industrial Technology Research Institute (ITRI)
- 2.5-month business trip in silicon valley. Look for potential business & research partners and also promote AI products developed in ITRI
- Successfully build up a long-term relationship with DirectDerm, a startup for telemedicine in dermatology.
Artificial Intelligence Nanodegree and Specializations, 05/2017 - 02/2018
- Online Course, Udacity
- Learn to build up deep learning applications with python, tensorflow and keras
- 4 Projects:
  1. Sing Language Recognizer with Hidden Markov Model
  2. Facial Keypoints Detection with Convolution Neural Network
  3. Machine Translation (French vs. English) with Recurrent Neural Network
  4. End-to-end Speech Recognizer with Recurrent Neural Network
- AIND Certificate (click to view)

Deep Learning Nanodegree Foundation, 01/2017 - 05/2017
- Online Course, Udacity
- Learn the fundamentals of deep learning, including convolution neural network, recurrent neural network, generative adversarial network.
- 3 Projects:
  1. Image Classification with Python and Tensorflow
  2. TV Scripts generation with Recurrent Neural Network
  3. Human Face Generation with Generative Adversarial Network
- DLND Certificate (click to view)
IDEO U Design Thinking Courses, 06/2016 - 10/2016
- Online Course, IDEO
- Learn the mindset of design thinking and the methods to drive innovation
- Courses:
  1. Insights for Innovation Certificate (click to view)
  2. Storytelling for Influence Certificate (click to view)
  3. From Ideas to Action Certificate (click to view)
  4. Leading for Creativity Certificate (click to view)
- Selected to be an alumni coach after taking all these four courses.

Honors & Awards

Outstanding Youth Award, 2014
- Awarded by the director of college of medicine in National Taiwan University
- Selected to be one of the 7 outstanding youths in college of medicine due to the achievements both in clinics and research.
Altruism Award, 2012
- Awarded by the director of college of medicine in National Taiwan University
- Stand out from 30 classmates in terms of the leadership and the contribution to making learning environment better in class.