• Homepage
  • About Me

アインクラッド

Xiang Feng | 冯相

Education

  • Zhejiang University

    M.Eng. in CS 09/2022 - Present

  • Zhejiang University

    B.Eng. in CS 09/2018 - 06/2022

Interests/Skills

  • Differentiable reconstruction and generation of appearance and shape
  • Differentiable computational photography
  • Differentiable simulation
  • Hardware Engineering

Publication

Representative paper is highlighted.

Learning Photometric Feature Transform for Free-form Object Scan

Xiang Feng, Kaizhang Kang, Fan Pei, Huakeng Ding, Jinjiang You, Ping Tan, Kun Zhou, Hongzhi Wu

(Under revision of TVCG) [arxiv]

Gaussian Splashing: Unified Particles for Versatile Motion Synthesis and Rendering

Yutao Feng*, Xiang Feng*, Yintong Shang, Ying Jiang, Chang Yu, Zeshun Zong, Tianjia Shao, Hongzhi Wu, Kun Zhou, Chenfanfu Jiang, Yin Yang

Arxiv 2024 [project page] [paper]
*: contributed equally

ElastoGen: 4D Generative Elastodynamics

Yutao Feng*, Yintong Shang*, Xiang Feng*, Lei Lan, Shandian Zhe, Tianjia Shao, Hongzhi Wu, Kun Zhou, Hao Su, Chenfanfu Jiang, Yin Yang

Arxiv 2024  [project page]
*: contributed equally

GS^3: Efficient Relighting with Triple Gaussian Splatting

Zoubin Bi*, Yixin Zeng*, Chong Zeng, Fan Pei, Xiang Feng, Kun Zhou, Hongzhi Wu

SIGGRAPH Asia 2024  [project page]
*: contributed equally

Real-time Acquisition and Reconstruction of Dynamic Volumes with Neural Structured Illumination

Yixin Zeng*, Zoubin Bi*, Mingrui Yin, Xiang Feng, Kun Zhou, Hongzhi Wu

CVPR 2024 [project page]
*: contributed equally

Differentiable Dynamic Visible-Light Tomography

Kaizhang Kang*, Zoubin Bi*, Xiang Feng, Yican Dong, Kun Zhou, Hongzhi Wu

SIGGRAPH Asia 2023 [project page]
*: contributed equally

Misc

Asuna: Vulkan-based Path Tracer

Asuna is a renderer base on vulkan ray tracing pipeline with NVIDIA card. It supports both rendering with gui and headless rendering. It implements multiple importance Monte Carlo sampling and fundamental materials.  [code]

Circuit-level Differentiable Hardware for Visual Computing

Used a hardware-level(FPGA) diffrentiable device to capture geometry and appearance of real-world objects. We leverage a Basler racer line scanner camera to capture and optimize the physical properties of the LED lighting module, which is controled by a Zynq xilinx FPGA.

Simplay: implemented physically-based animation tools and algorithms

Implemented dynamics of Rigid body/cloth/elastic body/shallow wave/character animation. Stable fluid/PIC/FLIP/APIC/2-way coupling MPM were implemented. SIGGRAPH Paper Fast Simulation of Mass-Spring Systems and Position-Based Surface Tension Flow were also implenmented.  [GAMES103/201,cloth,pbstf]