Follow-Your-Shape: Shape-Aware Image Editing via Trajectory-Guided Region Control
This addresses a specific bottleneck in image editing for users needing precise shape transformations while preserving background content.
The paper tackles the problem of large-scale shape transformations in image editing, where existing flow-based models often fail to achieve intended shape changes or degrade background quality. The proposed Follow-Your-Shape framework achieves superior editability and visual fidelity, particularly for shape replacement tasks, as demonstrated on their new ReShapeBench benchmark.
While recent flow-based image editing models demonstrate general-purpose capabilities across diverse tasks, they often struggle to specialize in challenging scenarios -- particularly those involving large-scale shape transformations. When performing such structural edits, these methods either fail to achieve the intended shape change or inadvertently alter non-target regions, resulting in degraded background quality. We propose Follow-Your-Shape, a training-free and mask-free framework that supports precise and controllable editing of object shapes while strictly preserving non-target content. Motivated by the divergence between inversion and editing trajectories, we compute a Trajectory Divergence Map (TDM) by comparing token-wise velocity differences between the inversion and denoising paths. The TDM enables precise localization of editable regions and guides a Scheduled KV Injection mechanism that ensures stable and faithful editing. To facilitate a rigorous evaluation, we introduce ReShapeBench, a new benchmark comprising 120 new images and enriched prompt pairs specifically curated for shape-aware editing. Experiments demonstrate that our method achieves superior editability and visual fidelity, particularly in tasks requiring large-scale shape replacement.