GeoEdit: Local Frames for Fast, Training-Free On-Manifold Editing in Diffusion Models

• URL: http://arxiv.org/abs/2604.24238v1
• Date: Mon, 27 Apr 2026 09:47:02 GMT
• Title: GeoEdit: Local Frames for Fast, Training-Free On-Manifold Editing in Diffusion Models
• Abstract summary: Training-free editing typically re-runs the full denoising trajectory for every edit strength, making iterative refinement expensive.<n>We instead edit near the data manifold, where small local updates can replace repeated re-synthesis.<n>Our algorithm constructs a tangent frame via small perturbations to the initial noise and alternates small tangent moves with diffusion-based projections.
• Score: 9.1810262671279
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion models are a leading paradigm for data generation, but training-free editing typically re-runs the full denoising trajectory for every edit strength, making iterative refinement expensive. To address this issue, we instead edit near the data manifold, where small local updates can replace repeated re-synthesis. To enable this, we estimate a local manifold tangent space directly from perturbed samples and prove that this sample-based estimator closely approximates the true tangent. Building on this guarantee, we devise a Jacobian-free algorithm that constructs a tangent frame via small perturbations to the initial noise and alternates small tangent moves with diffusion-based projections. Updates within this frame follow principled on-manifold directions while suppressing off-manifold drift, enabling fine-grained edits without full re-diffusion or additional training. Edit strength is controlled by the number of steps for rapid, continuous adjustments that preserve fidelity and plug into existing samplers. Empirically, the resulting tangent directions yield smooth, semantic unsupervised traversals and effective CLIP-guided optimization, demonstrating practical interactive continuous editing.

Related papers

• MLLM-Guided Semantic Correction for Text-to-Video Generation [66.7723669725808]
  We introduce a training-free, interpretable mid-generation correction framework that integrates multimodal large language model feedback.<n>We propose two key modules: a Semantic Assessment Supervisor that generates intermediate preview frames for semantic evaluations and deviation diagnostics, and a Semantic Modification Assistant that corrects semantic drift during inference.<n>Our method improves semantic alignment, visual fidelity, and temporal consistency without modifying model parameters.
  arXiv  Detail & Related papers  (2026-08-17T12:54:10Z)
• Consistent-Inversion: Reverse Consistency Guidance for Structure-Preserving Visual Editing [41.38183848746174]
  Consistent-Inversion is a training-free reverse consistency guidance framework for structure-preserving visual editing.<n>It improves background and structural fidelity under a unified SD3.5 protocol while maintaining target-prompt alignment.
  arXiv  Detail & Related papers  (2026-06-05T11:00:12Z)
• Semantic Granularity Navigation in Image Editing [4.8073791571301125]
  NaviEdit is a training-free inference-time controller for real-image editing.<n>It decouples edit progress from model scale through a strict self-consistency contract.<n>It treats scale as a control input and reallocates a fixed step budget toward semantically responsive intermediate scales.
  arXiv  Detail & Related papers  (2026-05-20T13:53:13Z)
• DirectEdit: Step-Level Accurate Inversion for Flow-Based Image Editing [51.56484100374058]
  We propose DirectEdit, a training-free editing method for pre-trained text-to-image (T2I) models.<n>DirectEdit eliminates the inherent reconstruction error without introducing additional neural function evaluations (NFEs)<n>Experiments demonstrate that DirectEdit achieves efficient and accurate image editing, delivering superior performance that outperforms state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-05-04T10:09:18Z)
• SteerFlow: Steering Rectified Flows for Faithful Inversion-Based Image Editing [9.343095346668665]
  SteerFlow is a model-agnostic editing framework with strong theoretical guarantees on source fidelity.<n>We introduce Trajectory Interpolation, which adaptively blends target-editing and source-reconstruction velocities to keep the editing trajectory anchored to the source.<n>Experiments on FLUX.1-dev and Stable Diffusion 3.5 Medium demonstrate that SteerFlow consistently achieves better editing quality than existing methods.
  arXiv  Detail & Related papers  (2026-04-02T07:26:13Z)
• FlowBypass: Rectified Flow Trajectory Bypass for Training-Free Image Editing [10.304374060580828]
  Training-free image editing has attracted increasing attention for its efficiency and independence from training data.<n>Previous attempts to address this issue typically employ backbone-specific feature manipulations, limiting general applicability.<n>We propose FlowBypass, a novel and analytical framework grounded in Rectified Flow that constructs a bypass directly connecting inversion and reconstruction trajectories.
  arXiv  Detail & Related papers  (2026-02-02T08:37:00Z)
• Adaptive Domain Shift in Diffusion Models for Cross-Modality Image Translation [35.54089670586124]
  Cross-modal image translation remains brittle and inefficient.<n>Standard diffusion approaches often rely on a single, global linear transfer between domains.<n>We embed domain-shift dynamics directly into the generative process.
  arXiv  Detail & Related papers  (2026-01-26T16:00:36Z)
• On Exact Editing of Flow-Based Diffusion Models [97.0633397035926]
  We propose Conditioned Velocity Correction (CVC) to reformulate flow-based editing as a distribution transformation problem driven by a known source prior.<n>CVC rethinks the role of velocity in inter-distribution transformation by introducing a dual-perspective velocity conversion mechanism.<n>We show that CVC consistently achieves superior fidelity, better semantic alignment, and more reliable editing behavior across diverse tasks.
  arXiv  Detail & Related papers  (2025-12-30T06:29:20Z)
• Edit-Based Flow Matching for Temporal Point Processes [51.33476564706644]
  temporal point processes (TPPs) are a fundamental tool for modeling event sequences in continuous time.<n>Recent non-autoregressive, diffusion-style models mitigate these issues by jointly interpolating between noise and data.<n>We introduce an Edit Flow process for TPPs that transports noise to data via insert, delete, and substitute edit operations.
  arXiv  Detail & Related papers  (2025-10-07T15:44:12Z)
• TAG:Tangential Amplifying Guidance for Hallucination-Resistant Diffusion Sampling [53.61290359948953]
  Tangential Amplifying Guidance (TAG) operates solely on trajectory signals without modifying the underlying diffusion model.<n>We formalize this guidance process by leveraging a first-order Taylor expansion.<n> TAG is a plug-and-play, architecture-agnostic module that improves diffusion sampling fidelity with minimal computational addition.
  arXiv  Detail & Related papers  (2025-10-06T06:53:29Z)
• FlowDirector: Training-Free Flow Steering for Precise Text-to-Video Editing [2.7248421583285265]
  FlowDirector is a novel inversion-free video editing framework.<n>Our framework models the editing process as a direct evolution in data space.<n>To achieve localized and controllable edits, we introduce an attention-guided masking mechanism.
  arXiv  Detail & Related papers  (2025-06-05T13:54:40Z)
• FlowAlign: Trajectory-Regularized, Inversion-Free Flow-based Image Editing [47.908940130654535]
  FlowAlign is an inversion-free flow-based framework for consistent image editing with optimal control-based trajectory control.<n>Our terminal point regularization is shown to balance semantic alignment with the edit prompt and structural consistency with the source image along the trajectory.<n>FlowAlign outperforms existing methods in both source preservation and editing controllability.
  arXiv  Detail & Related papers  (2025-05-29T06:33:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.