Projects

A VR network visualization system using voice commands and LLM-powered agents to transform natural language into Neo4j queries, enabling interactive graph exploration through categorical coloring, shape encoding, and dynamic layouts.

A real-time GPU-driven point cloud visualization system built in Unity, capable of rendering and interactively exploring large-scale LiDAR sensor data from autonomous vehicle datasets.

This project utilized the WiiFit Balance Board to investigate user balance disparities between playing first-person shooter (FPS) games in 2D mode and virtual reality (VR) environments.

Congratulations to the S14 champions T1!

In this RPG game, players take on the role of a stray cat and must navigate survival by completing storyline quests and interacting with characters and objects within the game world.

Unity tool that allows you to manage the selection history and customized categories dynamically.

Only one click! Easily convert your 2D image to a 3D model in Blender!

A research project aims to realtime display the physiological data during VR psychological therapy through Unity.

DappChef is a website similar to Leetcode, we aim to help people learn how to write a smart contract!

A official commemorative game for the 50th anniversary of the College of Engineering at NTHU.

This is a tool which helps artists or game developers to simply customize their bullet shooting system and visual effects!

This study explored using virtual scenarios to trigger drug cravings and physical responses. The research highlights the importance of tailoring scenarios to patients' recovery stages and contexts, offering insights for designing VR-based psychotherapy.

This is a step-by-step tutorial in order to help people build a login system using Firebase and Cocos Creator!

A commercial website for Sanyutang, where users can place orders for Taiwanese tea products.

We developed an EEG-based algorithm and interface to visualize arousal states in asymmetric multiplayer VR games in order to enhance social connection.

LLM explanations are often fluent but unreliable, making them difficult to verify. This project converts LLM outputs into structured Claim–Evidence–Assumption graphs, improving verification accuracy and reducing false acceptance in debugging tasks.

Built a Mondrian-style picture generator, blending coding and design to produce dynamic compositions inspired by Piet Mondrian's iconic artwork.

This is a research project aims to find out which visualization ways is more suitable to display user's concentration state during VR immersion.

This website helps people to better realize the information of published games, and provides some analysis for users to compare the games through different categories.

Use Deep Q-Networks (DQN) to train an autonomous agent to play a custom-built Chrome Dinosaur game in Python. A CNN-based model processes gameplay frames and obstacle information to perform real-time decision-making and obstacle avoidance.