List of Topics

We have explorations on following research directions :

Deep Learning Theory: meta-learning, domain adaptation, generalization, neural tangent kernel
Machine Learning Applications: people detection and tracking, OCR, document summary, text generalization
Reinforcement Learning: multi-task RL, multi-agent, distributional RL
Others: Transformer-based models, Graph Representation Learning

Ongoing Thesis Topics

The list of undergraduated thesis topics of this semester :

Apply wasserstein distance for improving policy gradient in navigation-based reinforcement learning
A study of generalization of meta-learning
Neural tangent kernel for deep equilibrim models
Learning Tabular data with transformer-based models
Vietnamese handwritten recognition with transformer
Transformer models as the universal embedding for mix-type input data
Seismic facies classification with unet model
Unnormal behaviour detection from crowds, a domain adaptation-based approach

The list of participated projects :

AI-based application for 3D-model restoration of Vietnamese and Korean cultural heritage (2021-2024). Funded by: Bilateral funded by Vietnam (MOST) and Korea (MSICT). Role: Main Investigator
AI-based system for integrating geophysics data and and detecting petroleum-potential (2020-2021). Funded by: Vietnam (MOST). Role: Main Investigator
Methods for improving Deeep CNN-Based Image Steganography (2020-2021). Funded by: VNU University of Engineering and Technology. Role: Principal Investigator
Research and development of application for counseling students in university using artificial intelligence (2017-2019). Funded by: Vietnamese Ministry of Education and Training. Role: Main Investigator
Analytical methods of satellite image processing for air pollution estimation in Vietnam (2013-2014). Funded by: Asia Research Center, Vietnam National University Hanoi. Role: Main Investigator
Recognition and Analysis of user context from multi-source data for Ambient Intelligent systems (2012-2015). Funded by: National Foundation for Science & Technology Development. Role: Main Investigator

Others :

Developing a realtime feed suggestion system based on social graph features (2019-2020). Gapo social network

Improve social-based product recommendations with transformer layers
Multi-agents learning with counter factual: A study in the case of long horizon sparse rewards
Weakly segmentation
Seismic data restoration with cycle-GAN
Han-nom OCR detecion based on differential binarization