Juan-ting Lin
Senior Computer Vision Researcher Engineer at Magic Leap
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolJuan-ting Lin is a Senior Computer Vision Researcher/Engineer with a decade of experience building practical vision systems, currently advancing spatial computing at Magic Leap in Zurich. Trained at ETH Zurich and with research roots at National Tsing Hua University and Academia Sinica, he blends rigorous academic methods with hands-on engineering for applications from robotic navigation to AR. His work spans deep learning for image enhancement (including hands-on improvements to an SRGAN TensorFlow implementation) to semantic segmentation and place recognition, reflecting both research and production sensibilities. At Magic Leap he progressed from researcher to senior engineer, evidencing impact on product-facing vision features. Comfortable teaching and mentoring, he has supported university courses in computer vision and signal processing, suggesting strong communication alongside technical depth. Colleagues would note his knack for translating cutting-edge papers into stable, debuggable code for real-world systems.
10 years of coding experience
6 years of employment as a software developer
Master's degree, Information Technology and Electrical Engineering, Master's degree, Information Technology and Electrical Engineering at ETH ZurichNational Tsing Hua University
Github Skills (12)
mask-rcnn10
faster-rcnn10
deep-learning10
tensorflow10
python10
generative-adversarial-network10
super-resolution10
modeling9
trainings9
evaluation9
eval9
mlops8
Programming languages (4)
Github contributions (5)
brade31919/SRGAN-tensorflow
Jul 2017 - Jan 2019
Tensorflow implementation of the SRGAN algorithm for single image super-resolution
Role in this project:
ML Engineer
ML EngineerContributions:21 commits, 1 PR, 17 pushes in 1 year 6 months
Contributions summary:Juan-ting primarily focused on modifying and enhancing the SRGAN model implementation in TensorFlow. They updated the main script for training, testing, and inference modes, addressing file paths and configurations. The user integrated new features such as an inference mode for super-resolution and modified the model definition in `lib/model.py`. In addition, they addressed bugs and made adjustments to the training configuration.
cyclegandeep-learningsingle-imageresolutiongenerative-adversarial-network
brade31919/radar_depth
Sep 2020 - Jan 2021
Source code of the IROS 2020 paper "Depth Estimation from Monocular Images and Sparse Radar Data"
Contributions:4 commits, 3 pushes, 1 branch in 4 months
radarstereodeep-learningsparsecomputer-vision
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.