resume
This is a description of the page. You can modify it in '_pages/cv.md'. You can also change or remove the top pdf download button.
Contact Information
| Name | Jihyeon Lee |
| Professional Title | Researcher / Software Engineer |
| jihyeon@cs.stanford.edu |
Professional Summary
Researcher at Google Research, Stanford CS Alum. Building AI for social good and robust decision-making.
Experience
-
2020 - Present Mountain View, CA
Software Engineer
Google Research
Building SKAI, an AI system for detecting damage after natural disasters for humanitarian organizations.
- Collaborating with United Nations World Food Programme and GiveDirectly.
-
2019 - 2019 Mountain View, CA
Software Engineering Intern
Google, Geo Insights Team
Designed and built pipeline to perform weakly supervised segmentation on aerial imagery, including training model and evaluating predictions. Developed novel method that soft-labels masks by synthesizing data augmentations.
- TensorFlow, Apache Beam, OpenCV, Python imaging libraries, machine learning at scale
-
2018 - 2019 Stanford, CA
Researcher
Stanford Sustain Lab
Developed a weakly supervised geovisual search pipeline that learns how to perform detection from classification labels and uses data augmentation + human-in-the-loop data distillation. Applied to locate brick kilns in Bangladesh to monitor environmental regulations.
-
Advised by Stefano Ermon, David Lobell, Marshall Burke, Steve Luby Mentored by Nina Brooks - Received Firestone Medal (only 1 in CS Department)
-
-
2018 - 2018 San Francisco, CA
Software Engineering Intern
Pinterest, Visual Search Team
Implemented data distillation to automatically augment training datasets from billions of unlabeled, user-generated imagery at little to no cost. Expanded dataset size by 200% and showed newly trained model performs better.
- Kleiner-Perkins Engineering Fellow
- TensorFlow, Hadoop/SQL, Pandas, Python imaging libraries
-
2017 - 2019 Stanford, CA
Researcher
Stanford Vision & Learning Lab
Investigated an active learning system that generates questions about images on social media and asks them to users to automatically collect data. Created a sequential attention-based model to handle noisy responses.
-
Advised by Fei-Fei Li, Michael Bernstein Mentored by Ranjay Krishna - Gave spotlight talk at CS 231N (Spring 2017)
- Presented at Stanford Human-Centered AI Symposium (Spring 2019)
-
-
2017 - 2017 Sunnyvale, CA
Computer Vision Software Intern
Matterport
Developed algorithm that creates 3D point cloud models from real-time scanning that is robust to color balance and lighting changes within a single scene.
- C++, OpenGL, OpenCV, Android
-
2016 - 2017 Stanford, CA
Researcher
Stanford HCI Lab
Developed a system that uncovers insights for creative experts (journalists, podcasters) from large amounts of unstructured visual and language data. Created an interactive archive for Douglas Engelbart’s work.
-
Advised by Maneesh Agrawala Mentored by Mitchell Gordon, Juho Kim - {“Publication”=>”UIST ‘17 (30th Annual ACM Symposium on User Interface Software and Technology)”}
-
Education
-
2019 - 2020 Stanford, CA
M.S.
Stanford University
Computer Science (AI & HCI Specializations)
- Siebel Scholar (full merit-based scholarship)
- {“Advisers”=>”Maneesh Agrawala and Stefano Ermon”}
-
2015 - 2019 Stanford, CA
B.S. with Honors
Stanford University
Computer Science (AI Specialization), Minor in Feminist, Gender, & Sexuality Studies
- Firestone Medal for Excellence in Undergraduate Research (top 10% of theses)