Yujia Chen (陈禹嘉)

Hey Amigo!

I'm a Machine Learning Engineer in Google Reseach, my current focus lies in applied research for real-time, on-device technologies in the fields of Computer Vision and Machine Learning. My expertise encompasses scene understanding, deep generative modeling, and representation learning. Before joining Google, I was an Applied Scientist at Amazon, where I developed innovative RGB-only computer vision algorithms for a range of products. This role followed my graduation from the MSCV program at Carnegie Mellon University's Robotics Institute, under the guidance of Prof. David Held and Kris Kitani.

My passion is in transitioning AI technology from academic theory to practical, real-world applications. Recently, I have been working on developing advanced vision technologies for mobile devices, including the Face Unlock feature on Google's Pixel 7 & 7 Pro (2022), Pixel Fold & 7a (Google I/O'23), and Pixel 8 and 8 Pro (2023). Before Google, I was part of a research team at Amazon, contributing to the development of vision-only technologies for the pioneering Just-Walk-Out (JWO) grocery store, also known as Amazon Go. Additionally, I have made significant contributions to Amazon Prime Video's Video Compliance System and the Virtual Product Placement feature, both of which are utilized globally.

Professional Experience

• Jun. 2022 - I joined Google Research as a Machine Learning Engineer in Seattle, Washington :D
• Jan. 2020 - I started a new adventure as an Applied Scientist at Amazon in the center of downtown Seattle!
• Sep. 2019 - Served as a Teaching Assistant for the "Introduction to Machine Learning" (10601) course at CMU, instructed by Prof. Matt Gormley.
• May 2019 - I started an amazing journey at Amazon as an Applied Scientist Intern, working on scene understanding & customer association.
• Dec. 2018 - Worked as a Graduate Assistant under Prof. David Held, concentrating on deep slope estimation and cloth part detection.
• May 2017 - I interned at National Laboratory of Pattern Recognition, Chinese Academy of Science, advised by Prof. Ran He.

Publications

1. Beyond First-order Tweedie: Solving Inverse Problems using Latent Diffusion
   Litu Rout, Yujia Chen, Abhishek Kumar, Constantine Caramanis, Sanjay Shakkottai, and Wen-Sheng Chu
   Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'24)
   [Project page]
2. Adversarial Occlusion-aware Face Detection (Oral)
   Yujia Chen, Lingxiao Song, Ran He
   International Conference on Biometrics Theory, Applications and Systems (BTAS'18), 2018
3. GM-Net: Learning Features with More Efficiency (Oral)
   Yujia Chen, Ce Li
   Asian Conference on Pattern Recognition (ACPR'17), 2017

Life

I have "twocatsandadog" (yes, that's my Wi-Fi password). The biggest boy cat, Noodle, only comes back home for food and sleep, and spends the rest of his life playing with the squirrels and birds in the wild. The sister cat, Pudding, is a little princess sleeping on the sofa all day. And the little brother, JoJo, is the famous White Golden Retriever in the neighborhood who likes to scramble in the mud!