人像重打光深度学习方法研究进展

Key Points

• This research aims to analyze advancements in portrait relighting techniques using deep learning. It focuses on the effectiveness and challenges of these methods in enhancing visual appeal.
• Summarized research background and application areas for portrait relighting.
• Reviewed existing portrait and lighting datasets.
• Classified and analyzed current deep learning-based relighting methods.
• Compared advantages and limitations of these methods across various scenarios.
• Deep learning techniques have become mainstream in effectively capturing lighting variations.
• Existing methods successfully enhance the aesthetic quality and realism of portrait images.
• Identified gaps in current research and potential future directions for algorithm improvement.