Linli Shi

Linli Shi

Master Student at Johns Hopkins University

👋 I'm Linli Shi 「史林立」

I am a master’s student at Johns Hopkins University, working with Prof. Ziyun (Claude) Wang in the SPIKE Lab. My current research interests include event-based vision, 3D vision, and robotics. To enable robots to reliably perceive and act in the fast-paced, dynamic real world is one of my research goals.

Before joining JHU, I received a dual degree in Computer Science: a BSc from the University of Birmingham and a BEng from Central South University. At the University of Birmingham, I worked with Prof. Hyung Jin Chang and Dr. Yihua Cheng on gaze estimation.

Outside research, I’m an SCA-certified barista and specialty coffee nerd ☕️. Happy to chat over a pour-over!

🔥 News

  • [2026-07] ✨ EVIS v1.0 is now available, bringing HDR support, configurable sensor noise, motion blur, anti-aliasing controls, and a PyTorch dataloader for streamlined downstream training. See our technical report for details.

📝 Publications

Isaac Sim Event Camera Plugin

EVIS: A Physics-Grounded Event Camera Plugin for NVIDIA Isaac Sim

Linli Shi, Ruijun Zhang, Ziyun Wang

arXiv preprint, 2026

[ Code] [📄 Paper]

  • A physics-grounded event camera simulation platform for NVIDIA Isaac Sim, bridging realistic event generation, controlled benchmarking, and downstream learning.
  • Supports event generation from linear intensity before gamma correction, realistic sensor noise, motion-blurred rendering, and anti-aliased rendering and warping.
  • Provides a PyTorch-native dataloader for multiple event representations, with evaluations on frame reconstruction, optical flow, and pose estimation.

🧩 Projects

RTDNet

RTDNet: A Recurrent Network using Temporal Difference for Video Gaze Estimation

Linli Shi, Hyung Jin Chang, Yihua Cheng

Undergraduate capstone project — awarded Outstanding Final Year Project, UoB

[ Code]

  • A recurrent network that alternates between spatial and temporal feature processing for high-precision video gaze estimation.
  • PDA Loss aligns predicted gaze trajectories with ground truth, moving beyond per-frame point consistency.

Backseat

Linli Shi, Abigail Pain, Hao Li, Yunfan Song, Ella Gardner, Xuanwei Shi, Joe Penny

[ Code]

  • A Spring Boot + Angular web application that augments Spotify with user-driven ratings, reviews, leaderboards, and social music discovery.

🎖 Honors

  • 2023-2024 Dean’s List, University of Birmingham
  • 2023, 2024 University of Birmingham Entrance Scholarship
  • ☕️ 2025.09 SCA Brewing Intermediate, Specialty Coffee Association

📖 Educations

  • 2025.01 - Now, Johns Hopkins University, MSE Electrical and Computer Engineering
  • 2023.09 - 2025.06, University of Birmingham, BSc Computer Science
  • 2021.09 - 2025.06, Central South University, BEng Computer Science

💻 Experience

  • 2025.01 - 2025.05, Teaching Assistant of 30205 Team Project AI @ UoB
  • 2023.12 - 2024.01, Software Engineer Intern @ hibretOne-CIC
  • 2022.09 - 2025.06, Technical Consultant @ Apple Lab (iOS Club), CSU
  • 2022.09 - 2025.06, Member @ Open Source Association, CSU
  • ☕️ 2025.10 - 2025.12, Barista @ Seesaw Coffee, Taiyuan