HirFormer: Dynamic High Resolution Transformer for Large-Scale Image Shadow Removal
Computer Vision and Pattern Recognition（2024）
关键词
High-resolution,Large-scale Image,Shadow Images,Large-scale Removal,Shadow Removal,High-resolution Images,Background Information,Structural Similarity,Model Performance,Validation Set,Stage 2,Batch Size,Input Image,Effective Removal,Large-scale Networks,Patch Size,Peak Signal-to-noise Ratio,Training Step,Transformer Model,Two-stage Model,Ground Truth Image,Vision Transformer,Image Deblurring,Refinement Network,Objective Metrics,Deep Learning-based Approaches,Shadow Regions,Frequency Domain Information,Small Patches,Background Regions
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要