🚀 My Projects
A collection of my academic, research, and hackathon projects. Machine Learning & AI-focused projects are listed first.
Xu Lab – Cryo-EM Denoising & Reconstruction
Mar 2025 – Present | Carnegie Mellon University
Developing a novel diffusion-model-based computer vision pipeline for denoising and particle picking in cryogenic electron microscopy (cryo-EM) images. Replacing conventional Noise2Noise approaches with target score matching for improved biological particle reconstruction. Also generated synthetic cryo-ET datasets using PolNet to support deep learning development.
NeuroMechatronics Lab – Neural Modeling
May 2025 – Present | Carnegie Mellon University
Implementing deep learning models (including RNNs) to optimize the mapping from simulated excitatory input to target motor neuron output. Enhances simulation efficiency and recruitment dynamics in motor neuron models.
Adaptive Neurostimulation in Simplified Cortical Network
Mar 2025 – May 2025 | Python, PyTorch
Simulated a Pyr–PV neural circuit and designed a biologically grounded reward function to enhance pyramidal neuron activity while suppressing PV interneuron overactivation. Applied layered adaptive reinforcement learning (Bayesian Optimization, Temporal Difference Learning, Thompson Sampling) to identify effective stimulation patterns.
Neural Recovery Prediction from Pre-Post Stroke MRI
Mar 2025 – May 2025 | Python, FSL, SimpleITK, PyTorch
Developed a pipeline for detecting lesion-affected and compensatory brain regions in MRI scans. Extracted spatial features and patient metadata to train Random Forest and Deep Neural Network regressors to predict recovery scores. Explored Graph Attention Networks for functional recovery prediction.
Traffic Collision Rate Prediction in Los Angeles
Mar 2023 – May 2023 | Python, PyTorch
Built predictive models incorporating k-means clustering and decision-tree classifiers to identify critical factors behind traffic collisions. Delivered a research paper showing the effectiveness of ML-based event prediction for urban safety.
Nano Lab – Stretchable Synaptic Transistors
Mar 2023 – May 2024 | University of Southern California
Contributed to projects on SiO₂-based transistors. Performed material characterization and electrical testing, including SEM analysis and pulse-response testing for artificial synapse plasticity. Co-authored an ACS Nano paper on strain-insensitive, air-stable stretchable carbon nanotube-based synaptic transistor arrays for neuromorphic computing.
Intuitively-Controlled Robotic Arm
Jan 2024 – May 2024 | Python, Jetson Xavier NX
Built an anthropomorphic robotic arm capable of autonomously detecting and grasping objects (e.g., a coffee cup). Designed wireless communication between a PC and Jetson Xavier NX, implemented computer vision algorithms, and mapped object positions into servo motor rotations.
Photonic Inverse Design of On-Chip Power Splitter
Apr 2024 | MATLAB
Simulated a 1-to-4 silicon waveguide splitter using 2D Finite Difference Frequency Domain method with perfectly matched layers. Optimized dielectric distribution for low-loss, uniform power splitting.
Smart Biosphere in IoT
Dec 2022 | Python, Raspberry Pi
Created an IoT system integrating open-source weather APIs and real-time sensor data to control LED brightness, display conditions on an LCD, and visualize water-level data on Grafana. Used MQTT for efficient IoT communication.
IEEE Hack IoT (2nd Place)
Apr 2023 | USC IEEE Hackathon
Collaborated with a team of 4 in a 24-hour hackathon. Developed an IoT-based smart bicycle detection system using multi-sensory data integration. Achieved 2nd place overall.