Chiyu Max Jiang
Chiyu 'Max' Jiang (蒋驰宇)

Ph.D student at UC Berkeley.
Deep Learning methods for 3D Geometries and Physical Systems.

About

I’m a fifth-year Ph.D student at UC Berkeley advised by Prof. Philip Marcus, affiliated with the Data Analytics group at NERSC, Lawrence Berkeley National Lab. My research interest is in Machine Learning and Deep Learning methodologies related to 3D geometry. In particular, I am interested in leveraging advances in 3D learning for applications in a variety of physical and engineering systems, examples include 3D scene reconstruction, omnidirectional image segmentation, climate pattern detection and aerodynamical shape optimization.

Prior to joining UC Berkeley, I got my Bachelor’s degree from Cornell University, as well as a joint degree from Zhejiang University.

News

  • New! Two papers accepted to CVPR 2020! Check out my papers on local implicit grid representations for 3D scenes, and adversarial texture optimization from RGB-D scans.
  • My recent paper on Deep Differentiable Simplex Layer has been accepted to ICCV 2019 conference!
  • I will be interning in Machine Perception @ Google Research in summer 2019 as a Ph.D research intern.
  • My work on Spherical CNNs on Unstructred Grids has been chosen for an oral presentation at the AGU (Americal Geophysics Union).
  • Two of my papers have been accepted to the ICLR 2019 conference!
  • I am interning this summer at the Data Analytics group in NERSC, Lawrence Berkeley National Labratory, working with Prabhat and Karthik Karshinath on new Deep Learning methodologies for Climate Science.
  • I am invited to visit Center for Nonlinear Studies at Los Alamos National Labratory, and to present my work on 3D deep learning.
  • I made an oral presentation of my work on Aerodynamics Optimization using Deep Learning at Physics Informed Machine Learning.

Experiences

  • [Su 19] Research Intern at Google AI - Perception:
    Research on deep 3D geometric representation and reconstructions.

  • [Su 18] Research Intern at Lawrence Berkeley National Lab:
    Research on Spherical CNNs on Unstructured Grids and applications towards computer vision and climate science (subsequent publication at ICLR 2019).

  • [Fa 17] Graduate Student Instructor:
    (CS294-73): Software Engineering for Scientific Computing.

Professional Service

Reviewer [ICCV’19, AAAI’20, CVPR’20, ECCV’20, NeurIPS’20]

Publications

MeshfreeFlowNet: A Physics-Constrained Deep Continuous Space-Time Super-Resolution Framework
In submission
Chiyu "Max" Jiang*, Soheil Esmaeilzadeh*, Kamyar Azizzadenesheli, Karthik Kashinath, Mustafa Mustafa, Hamdi Tchelepi, Philip Marcus, Prabhat, Anima Anandkumar
Local Implicit Grid Representations for 3D Scenes
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2020)
Chiyu "Max" Jiang, Avneesh Sud, Ameesh Makadia, Jingwei Huang, Matthias Niessner, Tom Funkhouser

Adversarial Texture Optimization from RGB-D Scans
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2020)
Jingwei Huang, Justus Thies, Angela Dai, Abhijit Kundu, Chiyu "Max" Jiang, Leonidas Guibas, Matthias Niessner, Tom Funkhouser

DDSL: Deep Differentiable Simplex Layer for Learning Geometric Signals
Proceedings of the IEEE International Conference on Computer Vision (2019)
Chiyu "Max" Jiang*, Dana Lansigan*, Philip Marcus, Matthias Niessner

Spherical CNNs on Unstructured Grids
International Conference on Learning Representations (2019)
Chiyu "Max" Jiang, Jingwei Huang, Karthik Kashinath, Prabhat, Philip Marcus, Matthias Niessner

Convolutional Neural Networks on non-uniform geometrical signals using Euclidean spectral transformation
International Conference on Learning Representations (2019)
Chiyu "Max" Jiang, Dequan Wang, Jingwei Huang, Philip Marcus, Matthias Niessner

Leveraging Bayesian Analysis To Improve Reduced Order Models
Journal of Computational Physics (2019): 280-297.
B.T. Nadiga, Chiyu Max Jiang, Daniel Livscu

Finding the optimal shape of the leading-and-trailing car of a high-speed train using design-by-morphing
Computational Mechanics (2017): 1-23.
Sahuck Oh, Chung-Hsiang Jiang, Chiyu "Max" Jiang, Philip Marcus

Hierarchical Detail Enhancing Mesh-Based Shape Generation with 3D Generative Adversarial Network
Chiyu "Max" Jiang, Philip Marcus

Other Select Projects

Morphing of Genus-Zero Shapes using Spherical Parameterization  
Chiyu "Max" Jiang, Philip Marcus