Showcase
Selected projects highlighting my work in 3D computer vision, generative AI, virtual reality, and real-world software systems.
Featured Projects
OptimusMesh
OptimusMesh compresses a point cloud into 16 sparse latent pivots, then generates compact mesh vertices and faces with two autoregressive decoders.
Crowded Train Breaker
Interverse Virtual Reality Challenge (IVRC 2025), Japan
Participated in IVRC 2025 with the project "Crowded Train Breaker", representing the Graduate School of Information Science and Technology, The University of Osaka. The project successfully advanced to the SEED STAGE.
Worked as part of an 8-member team, contributing to both software and hardware development for an interactive virtual reality system. The project was presented at the Virtual Reality Society of Japan Conference, demonstrating an immersive system inspired by real-world crowded train scenarios.
Development Period: April 2025 – December 2025
Android Applications
Photo Recover
Developed an Android utility application for recovering deleted photos and improving file restoration usability on mobile devices.
Technologies: Java (Android), storage handling, mobile UI/UX design
Focus: Recovery utilities, file management, and practical mobile applications
Smart Switch: Phone Clone & File Transfer
Developed an Android application for phone-to-phone data transfer, cloning, and file sharing with a focus on usability and reliable transfer workflows. The application achieved 1M+ downloads on Google Play.
Technologies: Java (Android), file transfer systems, mobile UI/UX design
Focus: Data migration, mobile utilities, and large-scale real-world deployment
Ways Navigation – Offline GPS & Maps
Developed an Android navigation application focused on offline GPS functionality, route guidance, and map-based utilities for real-world travel use. The application achieved 100K+ downloads on Google Play.
Technologies: Java (Android), Google Maps APIs, mobile UI/UX design
Focus: Offline navigation, scalable mobile systems, and real-world deployment