Qinbo Li

I am a research scientist at Meta. I am interested in Computer Graphics, Computer Vision, and more specifically, computational photography, 3D vision and view synthesis.

Before I joined Meta, I received my PhD. from Texas A&M University, advised by Dr. Yoonsuck Choe.

Email  /  CV  /  Google Scholar  /  Github  /  Linkedin

profile photo
Research
Consistent View Synthesis with Pose-Guided Diffusion Models
Hung-Yu Tseng, Qinbo Li, Changil Kim, Suhib Alsisan, Jia-Bin Huang, Johannes Kopf
CVPR 2023
Project page / Arxiv

We propose a pose-guided diffusion model to generate a consistent long-term video of novel views from a single image.

Emergence of Tool Construction and Tool Use Through Hierarchical Reinforcement Learning
Qinbo Li and Yoonsuck Choe
Artificial Intelligence in the Age of Neural Networks and Brain Computing, 2023
TBA

We propose to use visual, proprioceptive and kinesthetic inputs to learn to construct and use tools with hierarchical reinforcement learning.

Video Face Recognition with Audio-Visual Aggregation Network
Qinbo Li, Qing Wan, Sang-Heon Lee, Yoonsuck Choe
ICONIP, 2021 (Oral Presentation)
pdf

We introduce an Audio-Visual Aggregation Network (AVAN) to aggregate multiple facial and voice information to improve face recognition performance.

Synthesizing Light Field From a Single Image With Variable MPI and Two Network Fusion
Qinbo Li and Nima Kalantari
SIGGRAPH Asia, 2020
project page / pdf / video / code

Synthesize light fields from a single image on general scene. Introduced a light field dataset that includes indoor and outdoor scene.

Enhancing Privacy Through an Interactive on Demand Incremental Information Disclosure Interface Applying Privacy by Design to Record Linkage
Hye-Chung Kum, Eric D. Ragan, Gurudev Ilangovan, Mahin Ramezani, Qinbo Li, Cason Schmit
SOUPS 2019
pdf

Our research investigates approaches to enhance privacy while supporting legitimate access for human decision making in record linkage.

Dynamical Analysis of Recurrent Neural Circuits in Articulated Limb Controllers for Tool Use
Han Wang, Qinbo Li, Jaewook Yoo, Yoonsuck Choe
IJCNN 2016
pdf

we investigate the dynamics of a simple recurrent neural network used in a nontrivial articulated limb control task in a tool use domain.

Emergence of Tool Use in an Articulated Limb Controlled by Evolved Neural Circuits
Qinbo Li, Jaewook Yoo, Yoonsuck Choe
IJCNN 2015
pdf

we investigate how the capability to use tools can spontaneously emerge in a simulated environment with Neuroevolution algorithm.

Others

Construction and Use of Tools through Hierarchical Deep Reinforcement Learning
Qinbo Li and Yoonsuck Choe
IROS workshop 2021

Tool construction and use challenge: Tooling test rebooted
Yoonsuck Choe, Jaewook Yoo, and Qinbo Li
AAAI Workshop, 2015

Experience
Meta 2/7/2022 - Present
Research Scientist
Wyze Labs 1/4/2021 - 5/1/2021
Research Scientist Intern
Texas A&M University 9/1/2017 - 12/31/2020
Research Assistant
Teaching Assistant

Template from Jon Barron,
Last updated Feb 2023.