2016-Present
NTHU Vision Science Lab: Research Assistant
Supervised by Min Sun.
Lead a project about road scene segmentation adaptation.
This work is under review for ICCV 2017
2011-2016
Taiwan Semiconductor Manufacturing Company:
Senior Engineer
Supervised by Da-Wen Lin.
Participate in N10/N20 development.
2008-2010
National Taiwan University: Master's Degree
Supervised by Chi-Kuang Sun.
Major in Photonics and Optoelectronics.
2004-2008
National Taiwan University: Bachelor's Degree
Major in Electrical Engineering.
Bio. I am currently a research assiatant at VSLab of Prof. Min Sun at NTHU, Taiwan. My research focuses on Deep Learning and its applications in Computer Vision, especially for the perception system for autonomous driving cars.

In addition to AI, I am also interested in Data Science and feel excited about applying it to solve real world problems. Before I joined VSlab, I had been working in the R&D Division at Taiwan Semiconductor Manufacturing Company (TSMC) for five years. I was in charge of analyzing the data collected to diagnose the failure modes of manufacturing processes in 10/20 Nanometer Technology Project. To solve these industrial issues more efficiently, I aim to create a synergy between cutting-edge Data Science techniques and my existing domain expertise in Solid-State Physics.

To go deeper into these areas, I am going to pursue my second Master's Degree in Computer Science at the University of Illinois at Urbana-Champaign (UIUC) in 2017 Fall. See my CV for more details.

Publications

No More Discrimination: Cross City Adaptation of Road Scene Segmenters
We propose an unsupervised learning approach to adapt road scene segmenters across different cities. By utilizing Google Street View and its time-machine feature, we can collect unannotated images for each road scene at different times, so that the associated static-object priors can be extracted accordingly. By advancing a joint global and class-specific domain adversarial learning framework, adaptation of pre-trained segmenters to that city can be achieved without the need of any user annotation or interaction. We show that our method improves the performance of semantic segmentation in multiple cities across continents, while it performs favorably against state-of-the-art approaches requiring annotated training data.
Yi-Hsin Chen, Wei-Yu Chen, Yu-Ting Chen, Bo-Cheng Tsai, Yu-Chiang Frank Wang, Min Sun
Under Review for ICCV2017
Acoustic Velocity and Optical Index Birefringence in A-Plane ZnO Thin Films
The longitudinal acoustic velocity in the direction normal to the c-axis and the birefringence of refractive index of a-plane zinc oxide (ZnO) thin films were measured by ultrafast transient reflection measurement. Through Fourier analysis, the experimental data show that the acoustic velocity is about 6000 m/s and the measured refractive index near the band gap, which is higher than the previously reported results by 20 ∼ 30%.
Yi-Hsin Chen, Yu-Chieh Wen, Wei-Rein Liu, Wen-Feng Hsieh, Chi-Kuang Sun
CJP 2011
Femtosecond laser-ultrasonic investigation of plasmonic fields on the metal/gallium nitride interface
By using femtosecond laser-ultrasonic, we demonstrate an approach to study the surface plasmon field optically excited in the interface between metal and a semiconductor thin film. By femtosecond impulsive excitation on gallium–nitride GaN , different optical probe signals were observed when the impulse-excited nanoacoustic pulse propagated through the metal film and metal nanoslits. By analyzing the shape and temporal response of thus induced acousto-optical signals, our femtosecond laser-ultrasonic study not only reveals the plasmonic field distribution optically excited in the metal/substrate interface but also confirms that the penetration depth of surface plasmon field into the substrate agrees well with a simulation result.
Hung-Pin Chen, Yu-Chieh Wen, Yi-Hsin Chen, Cheng-Hua Tsai, Kuang-Li Lee, Pei-Kuen Wei, Jinn-Kong Sheu, Chi-Kuang Sun
Appl. Phys. Lett. 2010

Work Experiences

Full-Time RA in VSLab
Expertise in Caffe, Tensorflow, state-of-the-art Deep Learning based algorithms in Computer Vision (image detection/segmentation) and Domain Adaptation (domain adversarial training). Programming Experience in Python and MATLAB.
Senior Engineer at R&D Division
Five years of extensive work experience in problem-solving, stress-resistance, schedule management, and building collaborative relationships.