- 🎓 B.Sc. in Information Technology (AI) — VNU-HCM University of Science (GPA 3.77/4.0).
- 🧠 Focused on Multimodal LLMs (Vision + Language) & Android apps with on-device/connected AI.
- 🌱 Learning: Python, Kotlin/Compose, PyTorch, Docker for reproducibility.
Languages: Python • Kotlin
ML/DL: PyTorch • scikit-learn • NumPy • Pandas • Matplotlib/Seaborn • Pytorch
Practices: data preprocessing & feature engineering • training & evaluation • model implementation • prompt/LLM integration
Tools: Git • Docker • Jupyter • Pycharm • Android Studio • Gradle
-
Vision-Zephyr — Multimodal LLM for Visual Commonsense Reasoning
Python · PyTorch · CLIP · Zephyr LLM
▸ Implement & evaluate MLLM pipeline; multi-stage training (projector align → LoRA fine-tune); custom gating/fusion over CLIP layers.
Repo: https://github.com/baohuyvanba/Vision-Zephyr -
xWalletWise — Personal finance Android app with AI assistance
Kotlin · Jetpack Compose · Gemini API
▸ Expense tracking, budgeting, reports; integrates AI for insights & smart summaries.
Repo: https://github.com/baohuyvanba/AndroidSE-xWalletWise -
Deep-Learning — Implementations of classic DL models
Python · PyTorch
Repo: https://github.com/baohuyvanba/Deep-Learning -
Machine-Learning — From-scratch ML algorithms & experiments
Python · scikit-learn · NumPy
Repo: https://github.com/baohuyvanba/Machine-Learning
More projects
-
Virtual Try-On (CP-VTON re-implementation) — geometric matching & try-on modules; trained on 100k samples
Repo: https://github.com/baohuyvanba/hcmus_cvi_cp-vton_CSC16004 -
Predict Future Sales (Kaggle) — preprocessing, feature engineering, time-series forecasting
Repo: https://github.com/baohuyvanba/hcmus_dsp_finalproject_CSC17104 -
Pacman / Wumpus — search & logic agents (DFS/BFS/UCS/A*) and resolution-based reasoning
Repo: https://github.com/baohuyvanba/hcmus_aif_PacmanGame_CSC14003 -
Java Local Computer Remote — LAN remote-control tool (Java)
Repo: https://github.com/baohuyvanba/hcmus_mmt_JavaLocalRemote_CSC10008
- Internships/Fresher roles in AI/ML, Android, or Applied AI teams
- Research engineering (multimodal, VLM/MLLM, computer vision)
Reach me: [email protected] • LinkedIn