Pith. sign in

REVIEW 25 cited by

Direct Inversion: Boosting Diffusion-based Editing with 3 Lines of Code

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2310.01506 v2 pith:KUFIQF7O submitted 2023-10-02 cs.CV

Xuan Ju , Ailing Zeng , Yuxuan Bian , Shaoteng Liu , Qiang Xu This is my paper · ORCID
classification cs.CV
keywords editingimageinversionsourcebranchesdiffusiontargetperformance
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Text-guided diffusion models have revolutionized image generation and editing, offering exceptional realism and diversity. Specifically, in the context of diffusion-based editing, where a source image is edited according to a target prompt, the process commences by acquiring a noisy latent vector corresponding to the source image via the diffusion model. This vector is subsequently fed into separate source and target diffusion branches for editing. The accuracy of this inversion process significantly impacts the final editing outcome, influencing both essential content preservation of the source image and edit fidelity according to the target prompt. Prior inversion techniques aimed at finding a unified solution in both the source and target diffusion branches. However, our theoretical and empirical analyses reveal that disentangling these branches leads to a distinct separation of responsibilities for preserving essential content and ensuring edit fidelity. Building on this insight, we introduce "Direct Inversion," a novel technique achieving optimal performance of both branches with just three lines of code. To assess image editing performance, we present PIE-Bench, an editing benchmark with 700 images showcasing diverse scenes and editing types, accompanied by versatile annotations and comprehensive evaluation metrics. Compared to state-of-the-art optimization-based inversion techniques, our solution not only yields superior performance across 8 editing methods but also achieves nearly an order of speed-up.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 25 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. ReFlex: Text-Guided Editing of Real Images in Rectified Flow via Mid-Step Feature Extraction and Attention Adaptation

    cs.CV 2025-07 conditional novelty 7.0 of 10

    ReFlex edits real images with FLUX by extracting attention and residual features from a mid-step latent and adapting them during generation, improving text alignment while preserving structure.

  2. DNAEdit: Direct Noise Alignment for Text-Guided Rectified Flow Editing

    cs.CV 2025-06 conditional novelty 7.0 of 10

    Direct Noise Alignment iteratively moves a random Gaussian noise until the model's predicted velocity matches the straight-line velocity to the image, reducing inversion drift and giving the best reported fidelity-edi...

  3. Instruction-based Image Manipulation by Watching How Things Move

    cs.CV 2024-12 conditional novelty 7.0 of 10

    A diffusion editing model, InstructMove, is trained on video frame pairs annotated by MLLMs using spatial conditioning, enabling non-rigid edits and viewpoint changes.

  4. 4DHumanDiff: Direct Text-to-4DGS Generation for Consistent 360-Degree Dynamic Humans

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A diffusion model trained on 60,000 fitted 4D Gaussian Splatting human clips generates text-prompted, view-consistent dynamic humans directly in 4D, over 10x faster than video-first pipelines.

  5. FARI: Robust One-Step Inversion for Watermarking in Diffusion Models

    cs.CR 2026-07 accept novelty 6.0 of 10

    One-step adversarially LoRA-tuned inversion exploits low-curvature reverse trajectories to beat 50-step DDIM on watermark robustness after ~20 minutes of fine-tuning.

  6. Making Implicit Preservation Intent Explicit in Conversational Image Editing

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Conversational image editors fail to restore temporarily occluded content; ReSpec fixes this by explicitly selecting historical visual references and rewriting instructions to guide restoration.

  7. Discrete Noise Inversion for Next-scale Autoregressive Text-based Image Editing

    cs.CV 2025-09 conditional novelty 6.0 of 10

    VARIN uses a Location-aware Argmax Inversion pseudo-inverse of Gumbel-max sampling to extract editable discrete noises, enabling training-free prompt-guided editing for visual autoregressive models.

  8. Audio-Guided Visual Editing with Complex Multi-Modal Prompts

    cs.CV 2025-08 conditional novelty 6.0 of 10

    A training-free framework that maps audio embeddings into Stable Diffusion's text space and fuses multiple audio/text prompts via per-patch residual noise selection, outperforming text-only editors on new audio-visual...

  9. Trade-offs in Image Generation: How Do Different Dimensions Interact?

    cs.CV 2025-07 conditional novelty 6.0 of 10

    A new benchmark and VLM-as-judge metric map trade-offs among ten image-generation dimensions across 14 models, with a visualization called DTM.

  10. CPAM: Context-Preserving Adaptive Manipulation for Zero-Shot Real Image Editing

    cs.CV 2025-06 unverdicted novelty 6.0 of 10

    CPAM proposes a context-preserving adaptive manipulation method for zero-shot real image editing in diffusion models via preservation adaptation and localized extraction modules, outperforming prior techniques on a ne...

  11. AttentionDrag: Exploiting Latent Correlation Knowledge in Pre-trained Diffusion Models for Image Editing

    cs.CV 2025-06 conditional novelty 6.0 of 10

    AttentionDrag is a one-step, training-free drag-editing method that uses diffusion self-attention to move regions, generate masks, and fill gaps.

  12. Balancing Preservation and Modification: A Region and Semantic Aware Metric for Instruction-Based Image Editing

    cs.GR 2025-06 conditional novelty 6.0 of 10

    A region and semantic aware metric for instruction-based image editing, built from LLM parsing plus detection, segmentation, and CLIP directional similarity, reports the highest human alignment among compared metrics.

  13. RefEdit: A Benchmark and Method for Improving Instruction-based Image Editing Model on Referring Expressions

    cs.CV 2025-06 conditional novelty 6.0 of 10

    RefEdit-Bench measures referring-expression image editing; the RefEdit model, trained on 20K synthetic triplets, reports state-of-the-art results over million-scale baselines.

  14. Multi-turn Consistent Image Editing

    cs.CV 2025-05 conditional novelty 6.0 of 10

    The authors propose a FLUX-based multi-turn editing pipeline combining dual-objective LQR guidance and adaptive attention masking that maintains image identity over successive text edits, with the best FID score in fo...

  15. X-Fusion: Introducing New Modality to Frozen Large Language Models

    cs.CV 2025-04 conditional novelty 6.0 of 10

    X-Fusion adds a trainable vision tower to a frozen LLM, enabling both image captioning and text-to-image generation while keeping the LLM's language ability intact.

  16. Unpaired Multi-Domain Histopathology Virtual Staining using Dual Path Prompted Inversion

    eess.IV 2024-12 reject novelty 6.0 of 10

    A dual-path prompt inversion method performs unpaired virtual staining by matching a structural inversion trajectory and a style reference trajectory in a pre-trained diffusion model.

  17. Steering Rectified Flow Models in the Vector Field for Controlled Image Generation

    cs.CV 2024-11 conditional novelty 6.0 of 10

    FlowChef enables training-free, inversion-free, backprop-free controlled generation for rectified flow models by replacing the gradient through the model with the direct loss gradient on the estimated clean image.

  18. One Diffusion to Generate Them All

    cs.CV 2024-11 conditional novelty 6.0 of 10

    OneDiffusion shows that a single 2.8B-parameter diffusion model, trained by treating all tasks as frame sequences with varying noise scales, can handle image generation and image understanding tasks bidirectionally.

  19. Latent Space Disentanglement in Diffusion Transformers Enables Precise Zero-shot Semantic Editing

    cs.CV 2024-11 conditional novelty 6.0 of 10

    Diffusion transformer latent spaces are shown to be semantically disentangled, and prompt-difference directions plus a score-distillation step enable zero-shot fine-grained image editing.

  20. Diffusion Image Editing via Asynchronous Token Decoding

    cs.CV 2026-08 conditional novelty 5.0 of 10

    Token-wise surprisal selects editable regions, source key/value memory and hard projection preserve the rest, yielding record background-preservation scores on PIE-Bench without masks or fine-tuning.

  21. Inverse-and-Edit: Effective and Fast Image Editing by Cycle Consistency Models

    cs.CV 2025-06 conditional novelty 5.0 of 10

    A cycle-consistency reconstruction loss on the forward consistency model improves 4-step image inversion and editing, closing most of the quality gap to full-step diffusion editing.

  22. ByteMorph: Benchmarking Instruction-Guided Image Editing with Non-Rigid Motions

    cs.CV 2025-06 conditional novelty 5.0 of 10

    A released 6.4 million pair dataset and 613 sample benchmark for instruction-guided image editing of non-rigid motions, plus a Flux.1-dev based baseline that outperforms open-source methods on the new benchmark.

  23. FlowAlign: Trajectory-Regularized, Inversion-Free Flow-based Image Editing

    cs.CV 2025-05 conditional novelty 5.0 of 10

    FlowAlign adds a terminal-point source-similarity regularization to inversion-free flow-based editing, improving structural consistency while maintaining semantic alignment.

  24. Selftok: Discrete Visual Tokens of Autoregression, by Diffusion, and for Reasoning

    cs.CV 2025-05 reject novelty 5.0 of 10

    Selftok encodes images as diffusion-time-indexed discrete tokens, enabling a pure autoregressive VLM and visual RL with strong GenEval and DPG scores, though its claim that spatial tokens cannot support RL is not proven.

  25. DCI: Dual-Conditional Inversion for Boosting Diffusion-Based Image Editing

    cs.CV 2025-06 reject novelty 4.0 of 10

    DCI combines reference-guided noise correction with fixed-point latent refinement and reports state-of-the-art reconstruction and editing metrics on PIE-Bench.

Pith tools