Papers
Exploring Multispectral Information for the Detection of Marine Plastic Litter: Controlled Tank and Field Experiments A practical multispectral UAV framework for marine litter detection that combines controlled-tank band analysis, a compact RG+NIR input design, and super-resolution-assisted evaluation on coastal UAV imagery. Awarded Third Prize in the OCEANS 2026 Sanya Student Poster Competition.
Featured on IEEE Xplore, DOI Rose-Mamba-YOLO: an enhanced framework for efficient and accurate greenhouse rose monitoring A lightweight deep learning framework that enhances greenhouse rose monitoring by integrating Mamba-based feature modeling with YOLOv10 detection.
Featured on Frontiers, DOI Mamba-based super-resolution and semi-supervised YOLOv10 for freshwater mussel detection using acoustic video camera: A case study at Lake Izunuma, Japan A novel framework that integrates Mamba-based super-resolution with semi-supervised YOLOv10 to improve freshwater mussel detection in turbid underwater environments using acoustic video cameras.
Featured on ScienceDirect, DOI A Robust Detection Framework for Intelligent Growth Monitoring in Greenhouse Rose Cultivation Introduces ROSE-MAMBA-YOLO, a hybrid framework combining YOLOv11 with Mamba-inspired state-space modeling for robust rose detection in UAV imagery, achieving superior accuracy in challenging greenhouse environments.
Featured on IEEE Xplore, DOI Seafloor debris detection using underwater images and deep learning-driven image restoration: A case study from Koh Tao, Thailand An underwater monitoring framework that integrates super-resolution reconstruction with an optimized YOLOv8 detector (SFD-YOLO) to enhance image quality and achieve state-of-the-art accuracy in seafloor debris detection.
Featured on ScienceDirect, DOI Underwater Sea Cucumber Detection Using Consumer-Grade Amphibious UAV and Deep-Learning Based Computer Vision Presents a novel monitoring system that integrates a consumer-grade amphibious UAV with Mamba-based super-resolution and an enhanced YOLOv10 instance segmentation model to efficiently detect and map Holothurians in coral reef ecosystems.
Featured on IEEE Xplore, DOI Smart UAV-assisted rose growth monitoring with improved YOLOv10 and Mamba restoration techniques An integrated framework that combines Mamba-based super-resolution (MambaIR) with an improved YOLOv10 detector (ROSE-YOLO) to enhance UAV imagery for accurate and scalable greenhouse rose growth monitoring.
Featured on ScienceDirect, DOI YOLOv10 and Mamba-Based Super-Resolution for Smart Rose Growth Monitoring Using UAV Imagery A UAV-based monitoring framework that integrates Mamba-driven super-resolution with YOLOv10 detection to enhance imagery quality and achieve accurate growth assessment of greenhouse roses.
Featured on IEEE Xplore, DOI Projects
Multispectral Marine Litter Detection How a practical question about coastal marine litter became my first international conference — from a lab tank, to the Rokudōji coast, to the OCEANS 2026 Sanya poster stage.
Featured on IEEE Xplore, DOI RAG Evaluation Internship Nine months in Konica Minolta’s DX team — my first long-term internship in Japan, and my first time working on an enterprise-scale generative AI project. Low-light Gesture Recognition My first lab research project: teaching a model to recognize hand gestures in the dark — and the question about information that stayed with me afterwards. JiaCaiLiShu — Smart Steel Pipe Counting A WeChat mini program that counts steel pipes from a single photo — a team project that taught me how an algorithm becomes something people can actually use.