Po-Yi Chen Email & Phone Number

Kirkland, Wa, Us

Mountain View, Ca, Us

Seattle, Wa, Us

AWS Supply Chain Management Team- Designed and implemented a back-end service for validating and calculating data using Java and Typescript.

Seattle, Wa, Us

FrontEnd-Experience Team- Owned IDE import organization project, developed features that organize import statements for an Amazon front-end language on all IDE that follows language server protocols. Reduces hundreds of lines in the codebase and shortens development time for front-end development.- Designed organization rules for auto-complete import statement and implemented by using Kotlin with new data structures on the language server.- Developed two custom IDE commands. One that organizes import statements within a file at once and the other one removes unused import statements through a quick-fix. Implemented using Kotlin by sending specific requests from the client that will be handled by the language server and responded with the ideal text edit on the file.

Taipei, Northern Taiwan, Tw

Vision & Learning Lab, Advisor: Prof. Yu-Chiang Frank WangResearch interest: Computer Vision, specifically in-depth estimation from 2D scenes.- Outperformed state-of-the-art methods by decreasing 17% error rate in unsupervised depth estimation. Achieved by developing a multi-task deep learning model for scene understanding. Implemented using Python and Tensorflow by utilizing disjoint segmentation datasets and stereo images.- Published a first-author conference paper at CVPR 2019 for the proposed multi-task model and its results in unsupervised depth estimation. Been selected by the conference as one of the oral presentations.

Taipei, Northern Taiwan, Tw

Course EE4033: Algorithms, Instructor: Prof. Yu-Chiang Frank Wang

Taipei, Northern Taiwan, Tw

Course CommE5052: Deep Learning for Computer Vision, Instructor: Prof. Yu-Chiang Frank Wang

Taipei, Northern Taiwan, Tw

Communication and Multimedia Lab, Advisor: Prof. Yung-Yu Chuang- Implemented a Gaussian denoising model that utilizes residual learning. The model was able to remove different Gaussian noises from images. - Challenged a CVPR 2018 workshop topic about road extraction from aerial images by implementing and modifying a dilated residual network.

Taipei City, Taipei City, Tw

Multimedia and Machine Learning Lab, Advisor: Prof. Yu-Chiang Frank Wang- Researched and studied state-of-the-art deep learning models that could predict the 3D morphable models for human faces from 2D images.

Master'S Degree, Computer Science

Bachelor'S Degree, Computer Science