Yiran Zhang

I am currently a third-year Ph.D. in Computer Science at Purdue University under the supervision of Dr. Aniket Bera. Before this, I was graduated at Northwestern University with a CS Master Degree and was fortunate to work with Dr. Emma Alexander, Dr. Diego Klabjan, and Dr. Ian Horswill. I also received my B.S. in Computer Science and B.A. in Visual Fine Art double degrees from Union College. My research passion lies in utilizing different learning techniques to create innovative solutions for computer graphics, VR/AR, and locomotion problems.

Email  /  CV  /  Scholar  /  Twitter  /  Github

profile photo

Publications

XXXX Geometry Completion XXXX XXXX XXXX XXXX
Yiran Zhang, Hrishikesh Viswanath, Sainan Liu, Aniket Bera,
CVPR, 2026
Source Code and Dataset (Public Soon...)

We evaluate 3D generation SOTA model's ability to solve partial 2D to 3D completion task where we design a novel algorithm for views extraction on 3D fracture data for model tuning and evaluate the result geometry through various 2D and 3D distribution metrics to showcase its potential.

HCVR Scene Generation: High Compatibility Virtual Reality Environment Generation for Extended Redirected Walking
Yiran Zhang, Xingpeng Sun, Aniket Bera
Proceeding to IEEE VR, 2026
Source Code

HCVR virtual scene generation framework optimized for VR navigation and compatible with alignment-based redirected walking controller to provide user a unique natural walking interface.

TrustNavGPT: Modeling Uncertainty to Improve Trustworthiness of Audio-Guided LLM-Based Robot Navigation
Xingpeng Sun, Yiran Zhang, Xindi Tang, Amrit Singh Bedi, Aniket Bera
IROS, 2024
arXiv

An LLM-based audio-guided navigation agent that uses affective cues in spoken communication—elements to assess the trustworthiness of human commands.

Working Papers

2D Fractures to 3D Geometry Generation
Yiran Zhang, Hrishikesh Viswanath, Sainan Liu, Aniket Bera,
ECCV, 2026
Source Code (Public Soon...)

We extend our previous paper's fracture dataset to include fractured data from large complete 3D dataset with our decomposing algorithm. With larger dataset, our model can generate high fidelity 3D geometry model that captures the features appear in the partial 2D conditions. We also develop a new metrics to evaluate the feature similarity between the partial 2D data and the 3D result geometry.

Teaching and Mentoring

  • Teaching Assistant, COMP_SCI 182 Foundations of Computer Science, Purdue University, Spring 2025
  • Teaching Assistant, COMP_SCI 180 Programming I, Purdue University, Fall 2023-Fall 2025
  • Teaching Assistant, COMP_SCI 396 Advanced Computer Graphics Seminar: VR/AR Systems, Northwestern University, Summer 2023
  • Teaching Assistant, COMP_SCI 351-2 Advanced Computer Graphics, Northwestern University, Winter 2023
  • Teaching Assistant, COMP_SCI 327 Generative Methods, Northwestern University, Fall 2022

  • © Yiran Zhang | Last update: Sep. 19, 2025


    This website template is from Dr. Jon Barron. Thanks for his open-source spirit.