Building Human-aware Embodied Agent Systems

EEG & Multimodal Sensing → State Modeling → Decision → Robot Action

English · 简体中文

🌐 Personal Website

I’m an MEng student in Artificial Intelligence at UCLA, expecting to graduate in December 2027, with a computer science degree from Beijing Jiaotong University. I’m interested in embodied AI and robot learning, particularly how perception and learning methods connect to real systems.

My earlier work focused on EEG-based intention and emotion recognition, including model training, evaluation, and connecting predictions to a vehicle-control interface. During my internship at EBkernel, I worked on robot integrations involving EEG task events, facial expressions, and EMG gestures, as well as π0.5/OpenPI training and deployment on ALOHA. In my current internship at Philo Homes, I work on 3D reconstruction from residential videos, adapting research code, converting camera and scene data, and checking outputs.

Alongside this, I build web applications, backend services, and database projects. These give me experience with APIs, data storage, and deployment. Together, these projects have broadened my experience beyond model training to the data and software surrounding it—a foundation I want to build on in robot learning.

• EEG → Intention Decoding → Decision → STM32 Vehicle Control
• Facial Emotion → Decision → Safety-Gated Astribot Action
• EMG Gesture → Decision → ZsiBot Robot Control
• EEG Task Events → Human Confirmation → Robot Task Request
• Images + Robot State + Language Instruction → π0.5 via OpenPI → ALOHA Dual-Arm Action

My research interests include cross-subject EEG generalization, multimodal fusion across EEG, vision, language, and audio, as well as behavior-driven human state modeling beyond traditional BCI settings. I am also interested in deploying real-time AI for human-aware embodied agent systems.

• Multifractal + Graph-based + Transformer-based model
• Cross-subject generalization on SEED-VII dataset
• Focus on robustness and generalization
• Paper: Local-Global Feature Fusion for Subject-Independent EEG Emotion Recognition
• Accepted for Oral Presentation at IEEE EMBC 2026

Demo video: Watch the internship demo

• Multimodal human-state sensing with facial emotion recognition, EEG signals, and BCI paradigms
• Built a DeepFace-based emotion recognition prototype for real-time robot interaction — notes
• Explored SSVEP-based robot control and motor-imagery classification with EEGNet
• Connected EEG task events, human confirmation, and robot actions; added callback retries and logs to trace delivery failures
• π0.5 VLA model fine-tuning on AgileX Aloha for bimanual manipulation — notes
• OpenPI policy serving and dual-arm Piper inference with failure recovery — notes
• EMG gesture recognition controlling ZsiBot ZSL-1W wheeled-legged robot — notes

• 42-subject EEG-Audio-Video dataset with leakage-free splits; built complete unimodal baselines and late fusion achieving 0.5729 accuracy

• End-to-end EEG intention recognition system
• Model training (LSTM / SVM / etc.) + real-time control
• Integrated with embedded system (STM32 + Bluetooth)
• Honored as The 13th Cloud Programming World Cup - first prize

• Docker-first personal website with Caddy reverse proxy and WordPress
• Added a private FastAPI assistant for project-document search and server-status checks, with saved conversations and task history
• Notes: Amor Fati AI Infrastructure

• Programming: Python, C/C++, Java, JavaScript, SQL

• Machine Learning & Data: PyTorch, scikit-learn, NumPy, Pandas, OpenCV

• Web Development & Databases: FastAPI, Flask, LangGraph, MySQL, PostgreSQL, WordPress, Three.js

• Robotics: ROS, OpenPI

• Development Tools: Linux, Git, Docker, Bash, Nginx

I am open to collaboration and discussion on multimodal agent systems, embodied AI, EEG and vision fusion, human state modeling, scalable agent architectures, and real-world deployment of intelligent systems.

• Email: daniel.zhengzhou@gmail.com
• LinkedIn: www.linkedin.com/in/zheng-zhou-cs

Always learning to balance.