Chia-Hsin Chan
Principal Engineer, Video at Johnson Health Tech. Co., Ltd.- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
-
Chinese Native or bilingual proficiency
-
English Limited working proficiency
-
Japanese Elementary proficiency
Topline Score
Bio
Experience
-
Johnson Health Tech. Co., Ltd.
-
Taiwan
-
Wellness and Fitness Services
-
300 - 400 Employee
-
Principal Engineer, Video
-
Sep 2023 - Present
#Responsibility 1. Establish an AWS-based VOD and Live streaming platform 2. Video streaming performance and cost optimization #Responsibility 1. Establish an AWS-based VOD and Live streaming platform 2. Video streaming performance and cost optimization
-
-
-
DeepHow
-
United States
-
Software Development
-
1 - 100 Employee
-
Senior AI/ML Researcher
-
Mar 2022 - Apr 2023
# Responsibility 1. Solve video transcoding issues and improve performance 2. Study new technologies and open source projects for AI and video transcoding # Achievements 1. Video transcoding speed-ups - On AWS: saved 65~79% and 45~60% for video editing and upload time - On Ali-Cloud: saved 3~27% for video upload time - Ensure video quality with subjective quality assessment 2. Transcoding-free online video editor POC - Nginx-based video-on-demand service (Nginx-vod-module) via HLS - Perform erase, mute, voice-over editing on video with response time in seconds - Auto re-align captions for the edited video - Produce downloadable file with captions # Technology stacks - Python (Flask, FastAPI), Node.js, Docker - Google Cloud Platform, AWS, Ali-Cloud, Azure - Video transcoding (FFmpeg , AWS MediaConvert, Ali-Cloud ApsaraVideo Media Processing) - Knowledge of LLM, NLP, LangChain and chatGPT prompt engineering - Firebase, PostgreSQL, MongoDB - Git (GitLab), Jenkins Show less
-
-
-
Institute for Information Industry
-
Taiwan
-
Research Services
-
700 & Above Employee
-
Senior Engineer
-
Mar 2018 - Mar 2022
# Responsibility 1. Hardware selection and software development 2. Perform field survey and verify hardware/software setup for the traffic enforcement system 3. Writing proposals and reports/slides 4. Project leader (2019) 5. Intern interviewer # Achievements 1. Build a edge computing system for traffic event detection from scratch - Establish the architecture and code base - Develop 15+ traffic event detection/prediction functions based on CNN object detection: * RTSP live detection from CCTV camera or offline detection by MP4 files * Traffic law violation: auto generation of violence evidence photo/videos with LPR * Safety event: live detection with instant CMS warning * Live vehicle speed monitoring: by Smartmicro and U&U Engineering radars * Cloud event synchronization by MQTT * Solve the problem of multiple bounding boxes on the same object by 2nd MNS - Help Confucian Co., Taiwan's largest traffic law violation handling company, to import the automatic traffic law violation detection system - Design a camera view to real world mapping methodology based on geometric transformation 2. Backend API development for III EI-PaaS platform - Joint project with Advantech Co. - Develop backend APIs as information retrievers from Cloud Foundry - Implement micro-service communications by RabbitMQ - Improve API speed and reliability by threading, workflow re-design, and re-encapsulate Python requests library to have retry capability # Technology stacks - Python (OpenCV, Pytorch, Flask), Docker, MQTT, RabbitMQ - CNN object detection (YOLO v3/v4/v5, CenterNet), object tracking - Knowledge of deep learning trajectory prediction (LSTM, SocialGAN) - Speed radar software development - CloudFoundry - Git (GitLab), Jenkins, Scrum/Agile Show less
-
-
Education
-
National Chiao Tung University
Doctor of Philosophy - PhD, Computer Science -
National Chiao Tung University
Master of Science - MS, Computer Science -
University of Illinois at Urbana-Champaign
Bachelor of Engineering - BE, electrical and computer engineering -
National Chiao Tung University
Bachelor of Science - BS, Computer Science