Dr. LI Chengze

Assistant Professor
Yam Pak Charitable Foundation School of Computing and Information Sciences


Key Research Areas

  • 2D non‑photorealistic graphics and media analysis
  • Deep learning for image and video colorization, stylization, and synthesis
  • Animation, comics, and games (ACG) content understanding
  • Line art and sketch processing for assistive drawing tools
  • Human–computer interaction for creative media tools

 

 

 

 

Core Areas of Expertise

Dr. Li specializes in computer graphics and computer vision, with a focus on 2D non‑photorealistic media analysis. His expertise includes deep learning for image and video processing, computational photography, and animation, comics, and games (ACG) content understanding, emphasizing practical, production‑oriented solutions.

 

 Recent Publications

2026
  • See-through: Single-image layer decomposition for anime characters
2025
  • Advancing manga analysis: Comprehensive segmentation annotations for the Manga109 dataset
  • ColorDiffuser: Video colorization with pretrained text-to-image diffusion models
  • ChromaFlow: End-to-end flow matching for efficient video colorization
  • Instance-guided anime editing with a curated large-scale dataset
  • Screentone-preserved manga retargeting
  • Synchronized multi-frame diffusion for temporally consistent video stylization
  • Trajectory-guided anime video synthesis via effective motion learning
  • Cartoon animation outpainting with region-guided motion inference
  • Calligraphic guideline generation via contour-guided brush fitting
2024
  • A contrastive unified encoding framework for sticker style editing
  • Hyperstroke: A novel high-quality stroke representation for assistive artistic drawing
  • Appearance-preserved portrait-to-anime translation via proxy-guided domain adaptation
  • Separating shading and reflectance from cartoon illustrations
  • Shading-guided manga screening from reference
  • SKETCH2MANGA: Shaded manga screening from sketch with diffusion models
  • Body part segmentation of anime characters
 2023
  • Instance-guided cartoon editing with a large-scale dataset
  • Panel-page-aware comic genre understanding
  • AddCR: A data-driven cartoon remastering