CVPR 2026
Restore Text First, Enhance Image Later
Two-stage scene text image super-resolution with glyph structure guidance.
🔬Minxing Luo
Research Intern at Ming Team, Ant Group | Recent MSc Graduate
Ming Team, Ant Group
I am a recent MSc graduate from Nankai University, interested in generative vision, diffusion models, visual text generation, image editing, and text-aware image super-resolution. My work aims to make generated and enhanced imagery more structurally faithful and visually convincing.
Selected Publications
Each thumbnail is an architecture figure from the corresponding paper.
ICCV Workshops 2025
Beyond Flat Text
Dual self-inherited guidance for visual text generation.
arXiv 2024
RotationDrag
Point-based image editing with rotated diffusion features.
Experience
Research Intern
Ming Team, Ant Group
Conducting research in generative AI.
Research Intern
Imaging Algorithm Center, vivo
- Led the development of TIGER, a diffusion-based, text-aware image super-resolution framework accepted to CVPR 2026.
- Built real-world data pipelines and datasets for text image enhancement.
Education
MSc in Computer Science
Nankai University
Visual Computing and Intellectual Perception Lab.
BSc in Information Security
China University of Mining and Technology
Selected Honors
Interdisciplinary Contest in Modeling — Honorable Mention
COMAP ∙
May 2022
Honored for interdisciplinary modeling and problem-solving.
National Scholarship
Ministry of Education of China ∙
December 2021
Received in the 2019–2020 and 2020–2021 academic years; awarded to the top 0.2% of students nationwide.
Languages
Chinese — Native speaker
English — C1 (PTE Academic: 88)
English — C1 (PTE Academic: 88)