Profile

PhD graduate specializing in handwriting recognition for mathematical expressions, with strong experience in online stroke modeling and building deep learning systems using graph neural networks, Transformer, language models, etc.

Selected Projects

CROHME2023
- Competition on recognition of handwritten mathematical expressions. Paper, Datasets
  Keywords: Handwritten Mathematical Expressions datasets, Online and Offline
Contextual Reasoning for CROHME
- Boosting Math Recognition with Contextual Text. Paper, Datasets
  Keywords: Contextual Text for Mathematical Expressions, Vision Large Language Model
Math Recognition Systems
- Graph Modeling and Graph Neural Networks for Online Handwritten Math Recognition. Paper
  Keywords: Online Stroke Modeling, Graph Neural Networks
- Tree-structured Transformer for Online and Offline Handwritten Math Recognition. Paper
  Keywords: Transformer, Tree-based Language Model
Historical Document Analysis
- Mathematical Expression recognition for Leibniz Manuscripts. Paper and Datasets are coming soon.
  Keywords: Historical Mathematical Expression, Low Resource Documents

2022/09 - 2025/12: PhD Student
- Laboratoire des Sciences du Numérique de Nantes, LS2N
- Supervisor: Harold Mouchère
2023/06 - 2023/09: Visit Researcher
- DPRL, Rochester University of Technology, NY, USA
- Supervisor: Richard Zanibbi
2020/09 - 2021/08: Machine Learning Intern
- CAPACITÉS SAS, Supervisor: Patrick Le Callet
- Project cooperate with ByteDance. Paper
- Project cooperate with Tencent. Paper
2018/06 - 2018/09: Software Development Intern
- Huawei Suzhou Research Institute

Technical
- Deep Learning: Vision-Language Models, Graph Neural Networks, Transformer-based decoders, etc.
- Programming & Frameworks: Python, PyTorch, Optuna, HuggingFace, etc.
Research
- Experimental design and evaluation of ML models
- Data collection and annotation
- Scientific writing and peer-reviewed publications
Languages
- English (Fluent), French (Fluent), Chinese (Native)