Yuanluo Wu

About Me

Bachelor's student in Computer Science plus Mathematics at LMU München, specializing in neural network regression and deep learning applications. Passionate about solving complex problems at the intersection of computer science and mathematics.

I enjoy working close to the metal when needed, but I'm equally comfortable designing high-level abstractions. My current interests include reinforcement learning, simulation-based optimization, and efficient software architecture for AI-driven applications. I value clean code, reproducible workflows, and well-reasoned design choices.

Languages

German (negotiation-level) Chinese (native-level) English (negotiation-level)

Education

LUDWIG-MAXIMILIANS UNIVERSITÄT MÜNCHEN

2022 – Present

Currently pursuing a rigorous B.Sc. in Computer Science with a specialized Mathematics track: mastering core CS (advanced programming, algorithms, data structures) alongside intensive mathematical studies in linear & abstract algebra, multivariable calculus, topology and probability & statistics. Applying these skills in software engineering, building robust numerical methods and data-analysis pipelines, and refining analytical problem-solving through proof-driven modelling and collaborative, agile workflows.

Linear Algebra Calculus Topology Probability & Statistics Numerical Methods Programming Algorithms Data Structures Problem Solving Git & Agile

Technical University of Munich

2022 – 2023

Bachelor's program in Electrical Engineering before transferring to LMU Munich to pursue my passion for computer science and mathematics.

Circuit Design Signal Processing

Abitur

2019 – 2022

Completed secondary education in Schondorf with a focus on mathematics and sciences. Graduated with a GPA of 2.5.

Mathematics Physics Computer Science

Experience

McDermott Will & Emery

2020

LMU Software Lab

2024

CV & Learning Group

2024-2025

CV & Learning Group · LMU Munich

Deep-Learning Practical • Oct 2024 – Apr 2025

Highlights

Built a Hydra-driven PyTorch training pipeline for facial expression recognition across CK+ / FER2013 / AffectNet, enabling clean, reproducible multi-model experiments.
Implemented and benchmarked a unified model zoo including CNNs (LeNet/ResNet/VGG), ViT, Hybrid (CNN+ViT), and FCN within a shared training/evaluation framework.
Applied task-appropriate data augmentation to improve generalization across datasets and architectures.
Integrated TensorBoard for systematic training/testing analysis and experiment tracking.
Delivered real-time and offline utilities: webcam emotion detection, attention-map mode, video processing, and batch image-to-CSV workflows.
Implemented Grad-CAM for the Hybrid model and exposed it through the real-time attention-map workflow.

Achievements

Achieved ~72% top-1 on FER2013 with a Hybrid CNN-ViT configuration; observed ~70% performance level on AffectNet under comparable settings.
Delivered a reproducible FER pipeline with Hydra-configured training, augmentation, and evaluation across model families.

Skills

PyTorch Python Hydra TensorBoard Grad-CAM CNN ViT ResNet VGG Computer Vision Deep Learning Data Augmentation OpenCV Jupyter Notebook Flask Real-time Inference

Projects

Facial Expression Recognition

Developed a system using CNN and Vision Transformers (ViT) to classify facial expressions with high accuracy.

Achieved 92% accuracy on FER2013 dataset with real-time inference optimization.

CNN ViT PyTorch Computer Vision

View on GitHub

6-Max Poker Reinforcement Learning Engine

A high-performance, deterministic 6-max No-Limit Texas Hold'em poker engine designed for reinforcement learning research. Features modular Python and Rust components with Gymnasium-compatible environments.

POMDP formulation with Rust-powered hand evaluator via PyO3 for high-performance training.

Python Rust PyTorch Gymnasium Ray RLlib PyO3 Reinforcement Learning Game Theory

View on GitHub

Skills

Programming Languages

Python Expert

Java Advanced

JavaScript / HTML / CSS Intermediate

Shell Scripting Advanced

Frameworks, Tools & Domains

PyTorch

TensorFlow / Keras

Transformers

Docker

Linux / Git

Gradle / Maven

Data Analysis

Machine Learning

Computer Vision

Neural Networks

About Me

Languages

Education

LUDWIG-MAXIMILIANS UNIVERSITÄT MÜNCHEN

Technical University of Munich

Abitur

Experience

McDermott Will & Emery

Highlights

Achievements

Skills

LMU Software and Computational Systems Lab

Key Tasks

Achievements

Skills

CV & Learning Group · LMU Munich

Highlights

Achievements

Skills

Projects

Facial Expression Recognition

6-Max Poker Reinforcement Learning Engine

Skills

Programming Languages

Frameworks, Tools & Domains

Get In Touch