Advancing the Frontiers of Generative AI Research
W2GenAI Lab is a research laboratory dedicated to advancing the field of generative artificial intelligence. We focus on developing cutting-edge models and techniques for image generation, text-to-image synthesis, and multimodal AI systems.
Our mission is to push the boundaries of what's possible with generative AI while making these technologies accessible and useful for real-world applications.
- 🎨 AIGC Technology: Supporting the art creation industry and digital asset generation through advances in AIGC (Artificial Intelligence Generated Content) technology.
- 🌊 Visual Quality: Exploring visual challenges in photography to create images with superior quality.
🚀 LucidNFT — LR-Anchored Multi-Reward Preference Optimization for Generative Real-World Super-Resolution
🚀 UltraFlux — Data-Model Co-Design for High-quality Native 4K Text-to-Image Generation across Diverse Aspect Ratios
🚀 Kontext-Style — High-quality LoRA adapters for FLUX.1 Kontext; seamless style transfer and creative generation
🚀 LucidFlux — Caption-Free Universal Image Restoration with a Large-Scale Diffusion Transformer
🚀 Nano-Banana-Figure — LoRA for Qwen-Image-Edit (Real/Anime/Pet → Figure/Collectible)
Feel free to contact Tian Ye at tye610@connect.hkust-gz.edu.cn for cooperation, communication.