Yi (Jasper) Zhao
M.S. in Computer Vision student · Robotics Institute, Carnegie Mellon University
I am a master's student in the M.S. in Computer Vision (MSCV) program at the Carnegie Mellon University Robotics Institute (Aug 2026 – Dec 2027). I received my B.S. (Honours) in Data Science from Hong Kong Baptist University in June 2026.
Since April 2026, I have been working as a research assistant at the AirLab (Prof. Sebastian Scherer's group), CMU Robotics Institute, on foundation models for inertial perception — learning physically meaningful motion representations from raw IMU sequences, and building controllable generative models of inertial data to support robot state estimation.
Before moving into robotics, my research centered on medical image analysis: I developed segmentation-guided diffusion models for anatomy-preserving MRI generation (ICASSP 2026), deep-learning pipelines for material microstructure segmentation (AIPR 2025), and GAN-based super-resolution for diabetic retinopathy screening (ICDSE 2025).
Research interests: robot perception and state estimation, foundation models for sensing, generative models.
News
- Aug 2026Started the M.S. in Computer Vision program at the CMU Robotics Institute.
- Jul 2026We are hosting the TartanIMU Challenge: Multi-Platform Inertial Odometry at IROS 2026 — a Kaggle competition on estimating 3-D velocity from raw IMU with a single model across car, drone, quadruped, and human platforms. Official page · Kaggle.
- Apr 2026Joined the AirLab at CMU Robotics Institute as a research assistant, working on foundation models for inertial perception.
- Jan 2026Paper on segmentation-guided diffusion for MRI generation accepted at ICASSP 2026.
- Jun 2025Earned a Bronze Medal (ranked 60th / 1,136) in the Kaggle Featured Competition BYU — Locating Bacterial Flagellar Motors 2025, a 3D medical vision challenge on detecting bacterial flagellar motors in cryo-electron tomography volumes.
- 2025Papers accepted at AIPR 2025 (titanium alloy microstructure segmentation) and ICDSE 2025 (retinal image super-resolution).
Education
Carnegie Mellon University
Aug 2026 – Dec 2027M.S. in Computer Vision, Robotics Institute
Hong Kong Baptist University
Sep 2022 – Jun 2026B.S. (Honours) in Data Science
University of California, Berkeley
Jul 2024 – Aug 2024Summer Session — Probability & Statistics, Finance
Experience
-
AirLab, Robotics Institute, Carnegie Mellon University
Apr 2026 – PresentResearch Assistant · Prof. Sebastian Scherer's group
Learning-based state estimation and inertial perception for field robotics: foundation models that learn 3-D motion representations from raw 6-axis IMU sequences across heterogeneous platforms (vehicles, quadrupeds, drones, humans), and controllable generative modeling of inertial time series for robot state estimation.
-
Chinese Academy of Sciences
Jan 2025 – Feb 2025Research Assistant
Built an SRGAN-based super-resolution pipeline for diabetic retinopathy fundus images on the APTOS dataset — preprocessing, architecture design, perceptual-loss optimization, and two-stage adversarial training; the work led to a paper accepted at ICDSE 2025.
-
China United Network Communications (China Unicom)
Jun 2024 – Jul 2024Machine Learning Engineer Intern
Analyzed 10k+ user records and built PyTorch forecasting models for provincial mobile-data demand (96% accuracy), engineering temporal and geographic features to support promotion strategies.
Publications
Segmentation-Guided Mamba Dynamic Diffusion Model for Anatomy-Preserving MRI Generation
IEEE ICASSP 2026
SegMaDiff injects tumor segmentation masks at every denoising step of a Mamba-based dynamic diffusion model, achieving state-of-the-art FID on BraTS-2023/2024 while preserving anatomical structure, and enabling mask-only restoration of corrupted MRI slices.
Automated Microstructure Segmentation in Titanium Alloys Using Deep Learning Techniques
IEEE AIPR 2025
A deep-learning pipeline (VGG U-Net and SegFormer) for TC4 titanium alloy metallographic images, reaching 90% pixel accuracy and 0.7 mIoU on 1024×1024 images with a customized IWMID + Otsu preprocessing workflow.
Super-Resolution Image Generation for Diabetic Retinopathy Detection by SRGAN
ICDSE 2025
An SRGAN pipeline with a VGG-19 perceptual module and two-stage adversarial training for fundus image super-resolution on the APTOS dataset, improving texture fidelity for downstream retinopathy screening.