How a Medical AI company develops products

Report Title: How a Medical AI Company Develops Products

Speaker: Mr. Jiyong WANG (Export Algorithm Researcher)

Data: 27 September 2019 (Friday)

Time: 3:00 pm – 4:00 pm

Venue:  M08, Chu Hai College of Higher Education

Mr. Jiyong Wang received his Master of Information Technologies and Master of Philosophy in the medical image process from the University of Sydney in 2010 and 2012, respectively. He has seven-year experience in the medical image industry. Since 2012, he started working for Shanghai United-Imaging Healthcare Ltd. Co. and then from 2018 working for Ping An Health Technology Ltd. Co.

This presentation expects to give an insight into how a medical AI company develops products. And how traditional image filters can be used with convolutional neural networks. First, the segmentation of pulmonary vessels in chest CT images is introduced. Then, an idea of using hessian enhancement with convolutional neural networks is discussed.

For any inquiries, please feel free to contact Wendi Wang at 2972 7434.

Seminar: Digitalized Intelligent Evaluation Systems with Eye-tracking and Motion-tracking Technologies

On 23 August 2019, Prof. Hong Fu shared her valuable knowledge as a speaker during the seminar in Ping An Technology. The seminar is about the application of eye-tracking and motion-tracking technologies. Such technologies were integrated into a digitalized intelligent evaluation system and adopted in her research work.

Certificate of Appreciation

 

Video Dataset Based on Automated Cover Tests for Strabismus Evaluation

 

Video Dataset Based on Automated Cover Tests for Strabismus Evaluation

This video dataset based on the automated cover tests for strabismus evaluation (VD-ACTSE dataset) was developed to provide a useful dataset for the evaluation of strabismus. The video is acquired when the subject performs the automated cover tests. 24 samples were collected. The video was configured to have a resolution of 1280*720 pixels at a frame rate of 60 fps at a length of about 50s.

If you are interested in the dataset, please feel free to send an email entitled “VD-ACTSE dataset request” to hongfu@chuhai.edu.hk.

 

 

 

 

Figure 1 Some example frames of the video dataset

Light Field-based Face Spoofing Attack Dataset (LF-SAD)

The LF-SAD dataset was developed to provide a useful dataset for the evaluation of the light field-based face spoofing attack detection. It consists of three spoofing attack types—high definition printed photo, warped printed photo, and a high definition screen displayed photo. All photos in this database were captured by Lytro ILLUM light field camera at Chu Hai College of Higher Education.

Continue reading “Light Field-based Face Spoofing Attack Dataset (LF-SAD)”