The field of image editing is moving towards more efficient and effective methods, with a focus on preserving high-frequency features and achieving precise control over the editing process. Recent developments have introduced new frameworks and techniques that enable robust multi-turn editing, layered visual cues, and direct drag-based manipulation. These innovations have improved the quality and fidelity of edited images, allowing for more intuitive and interactive editing experiences. Notable papers include:

• Reversible Inversion, which achieves state-of-the-art performance with low computational overhead.
• FreqEdit, which enables stable editing across multiple iterations.
• MagicQuillV2, which introduces a layered composition paradigm for generative image editing.
• DirectDrag, which provides high-fidelity mask-free and prompt-free editing.