- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper introduces LayerEditor, a framework that jointly performs instruction-guided editing of a source image and generates an ordered stack of RGBA layers. It extends a pretrained image editor through joint RGB–Alpha token modeling and a separately trained Alpha decoder. Its Compositional Edit-Prior Distillation (C-EPD) queries a frozen editing teacher on student-visited states, supervising both the student’s direct canvas velocity and the decoded composition of its predicted layers. The authors construct LayerEditor-100K and evaluate on two layered-editing benchmarks, layer-count and instruction-control settings, and image-to-layer decomposition. The reported results show improved recomposition fidelity and layer quality over the evaluated baselines, with ablations supporting the complete C-EPD configuration.
Critical strengths
-
A useful framework for jointly editing images and generating controllable layers. This paper proposes LayerEditor, a system that generates an edited target image together with its ordered RGBA layer decomposition, supporting user-specified layer counts and layer-content descriptions. This structured output makes individual components accessible for subsequent manipulation, addressing a practical limitation of conventional editors that produce only flattened images.
-
A substantial paired dataset for instruction-guided layered editing. This paper introduces LayerEditor-100K, a dataset that covers text, object, and appearance edits with 2–10 layers per example, providing supervision for both editing outcomes and their underlying layer structures. The accompanying benchmarks evaluate recomposition fidelity, layer quality, and controllability. This data contribution is consequential because it addresses the scarcity of paired editing-and-layer annotations and provides a concrete foundation for training and evaluating models for this task.
Critical weaknesses
- The central claim of reusable layers is not directly validated in downstream manipulation. The quantitative evaluation primarily measures reconstruction, reference-layer agreement, and VLM-assessed quality in the original layouts. It does not systematically test whether moving, removing, resizing, or replacing generated layers preserves usable boundaries, recovers newly exposed content, and avoids residual artifacts.
- Limited validation of instruction-driven semantic grouping. Although LayerEditor supports layer-content descriptions and reports semantic-grouping scores in Table 2, the results do not clearly establish whether it can flexibly group multiple instances of the same category according to user intent. For example, can two birds be placed together in one layer when requested, or separated into different layers under an alternative instruction? A controlled comparison that varies the requested grouping while keeping the source image and intended edited appearance fixed would help distinguish instruction-driven grouping from category-based or instance-based decomposition.
- Limited validation of editing-capability retention. The five rubric dimensions are scored by a single Gemini-3.1-Pro evaluator without validation against human judgments (Appendix C.3). Figure 11 illustrates score anchors but does not establish agreement with human assessments. Although reference-based metrics provide complementary evidence, human blind evaluation and metric computation on an independent editing benchmark would strengthen conclusions about perceptual quality and editing-capability retention.
Questions for authors
Please review the weaknesses noted in my comments. I am open to increasing my score if the authors can properly address these concerns.
4 comments
2. Instruction-driven semantic grouping (W2)
LayerEditor uses text instructions to specify layer count and content. The Target layers section describes each layer in back-to-front order, with one entry per predicted RGB–Alpha pair. For example, L1: both people assigns both people to a single layer so they can be moved or scaled together. In contrast, L1: left person and L2: right person request separate layers, allowing either person to be moved, resized, or removed independently while leaving the other person unchanged. The layer count and descriptions must agree. Training and inference use the same four-section prompt, with an optional Prompt Enhancer assisting users who provide only a brief editing request (Appendix A.1, p. 14). Figures 15–16 (pp. 27–28) illustrate language-controlled grouping of people, buildings, and food items, followed by joint or independent manipulation. With the source image and intended edited appearance held fixed, LayerEditor achieves a semantic-grouping score of 6.88, compared with 5.91 for Seedream5-Pro (Joint) (Table 2, p. 8).
Table 2. Layer-count and semantic-grouping control with the source image and intended edited appearance held fixed (main paper, p. 8). Sem., Comp., and Alpha denote semantic-grouping, layer-composition, and Alpha-accuracy rubric scores on a 1–10 scale.
| Method | RGBA L1 ↓ | Alpha IoU ↑ | Sem. ↑ | Comp. ↑ | Alpha ↑ |
|---|---|---|---|---|---|
| Seedream5-Pro (Seq.) | 0.1184 | 0.4984 | 5.52 | 4.68 | 2.77 |
| Seedream5-Pro (Joint) | 0.1028 | 0.5703 | 5.91 | 5.15 | 4.28 |
| FR + QIL | 0.1625 | 0.2378 | 4.30 | 3.66 | 2.02 |
| FR + QIL-Control | 0.1175 | 0.4285 | 5.47 | 4.56 | 2.56 |
| FR + QIL-Decomp-FT | 0.1987 | 0.2554 | 5.65 | 5.47 | 1.36 |
| QIL-Edit-SFT | 0.1148 | 0.3096 | 5.10 | 4.57 | 1.60 |
| RevealLayer | 0.4096 | 0.2208 | 3.21 | 3.75 | 1.11 |
| LayerEditor | 0.0460 | 0.6196 | 6.88 | 6.28 | 6.98 |
Appendix D.1 (Table 8, p. 24) further varies editing instructions across 3,994 source documents, with a corresponding target image and reference RGBA stack for each request. Scores are averaged within documents before aggregation. Compared with Seedream5-Pro (Joint), LayerEditor improves edit consistency by 66.9% (3.32 → 5.54) and Composite SSIM by 10.8% (0.7914 → 0.8765), while reducing RGBA L1 by 38.9% (0.0563 → 0.0344). These improvements support LayerEditor’s ability to follow revised editing instructions while generating the corresponding layer organization.
Table 8. Instruction controllability across 3,994 source documents from Crello-Testset and ScaleEdit-Testset (Appendix D.1, p. 24). Editing instructions vary for each source image, and each prediction is evaluated against the corresponding target image and reference RGBA stack.
| Method | Composite SSIM ↑ | RGBA L1 ↓ | Alpha IoU ↑ | Edit. ↑ | Seg. ↑ | Comp. ↑ |
|---|---|---|---|---|---|---|
| Seedream5-Pro (Joint) | 0.7914 | 0.0563 | 0.5646 | 3.32 | 4.81 | 3.74 |
| FR + QIL | 0.4777 | 0.1435 | 0.1945 | 1.16 | 3.96 | 2.67 |
| FR + QIL-Control | 0.5827 | 0.0984 | 0.3502 | 1.45 | 5.11 | 3.87 |
| FR + QIL-Decomp-FT | 0.5972 | 0.1733 | 0.2123 | 1.41 | 4.68 | 4.34 |
| QIL-Edit-SFT | 0.7046 | 0.1978 | 0.2104 | 1.22 | 3.63 | 3.32 |
| LayerEditor | 0.8765 | 0.0344 | 0.5823 | 5.54 | 6.33 | 5.68 |
3. Human Evaluation of Editing Capability and Layer Reusability (W3)
Thank you for the constructive suggestion. ScaleEdit-Testset is built specifically from editing image pairs, with 1,000 held-out source documents and corresponding target layers (Section 3.4, p. 6; Appendix B.1–B.3, pp. 17–18). ScaleEdit-Testset tests whether requested edits remain correct after layer recomposition. In Table 1 (p. 6), LayerEditor improves Composite SSIM by 5.7% and reduces Composite FID by 33.8% relative to Seedream5-Pro (Joint). LayerEditor also achieves the highest rubric-based edit-consistency score, while the matched ablations confirm the editing improvements from C-EPD (Table 3, p. 9).
We added a method-blind human evaluation in which 50 evaluators with professional design backgrounds reviewed the same 500 cases sampled from both benchmarks. Evaluators scored editing, visual quality, layer content, composition, and reuse on a 1–10 scale (Section 4.6, p. 10; Appendix C.4, p. 23). LayerEditor leads all five dimensions in Table 5 (p. 10, shown above). Compared with w/o C-EPD, the mean human scores increase by 23.0% for edit consistency and 14.0% for visual quality, supporting the improvement trends observed with Gemini. Figures 11–12 (pp. 22–23) provide the Gemini rubric and scoring examples.
The user study also evaluates at least three consecutive rounds of layer reuse, with identical operation sequences and parameters across methods and no regeneration or repair. LayerEditor’s mean reuse score is 7.97, approximately 30.0% higher than w/o C-EPD and 65.7% higher than the strongest external baseline, FR + QIL-Control. Figure 8 (p. 10) illustrates independent layer operations, and Figures 15–16 (pp. 27–28) show different layer organizations and multi-round reuse. These results support both faithful initial editing and continued control of the generated layers, with usable recompositions after subsequent manipulation.
Table 1, ScaleEdit-Testset metrics (p. 6).
| Method | Composite SSIM ↑ | Composite FID ↓ | Edit Cons. ↑ | Visual Qual. ↑ |
|---|---|---|---|---|
| Seedream5-Pro (Seq.) | 0.7901 | 70.69 | 4.65 | 4.80 |
| Seedream5-Pro (Joint) | 0.8365 | 59.62 | 4.05 | 4.71 |
| FR + QIL | 0.6451 | 81.09 | 4.92 | 4.76 |
| FR + QIL-Control | 0.6606 | 84.07 | 4.70 | 5.08 |
| RevealLayer | 0.6464 | 75.74 | 4.28 | 4.08 |
| LayerEditor | 0.8839 | 39.48 | 5.96 | 6.58 |
Table 3. Ablations on Crello-Testset (p. 9).
| Model | Edit Cons. ↑ | Visual Qual. ↑ | Layer Seg. ↑ | RGBA L1 ↓ |
|---|---|---|---|---|
| LayerEditor | 6.14 | 6.35 | 6.83 | 0.0350 |
| w/o C-EPD | 4.67 | 5.12 | 5.01 | 0.0482 |
| w/o Direct OPD | 5.01 | 4.19 | 4.83 | 0.0513 |
| w/o Comp. OPD (Vanilla on-policy OPD) | 4.98 | 4.33 | 4.97 | 0.0492 |
| w/o Direct latent | 4.35 | 4.65 | 4.96 | 0.0642 |
| Base Model | 3.93 | 4.13 | 3.77 | 0.0735 |
Dear Reviewer nKre,
Thank you very much for your time and valuable feedback.
We understand that you may be busy. With the discussion deadline approaching and only one day remaining, we would appreciate it if you could review our responses and the revised manuscript.
Please let us know if any of your concerns remain unresolved. We would be very happy to address any further questions you may have.
Thank you for your attention.
Best regards, Authors of #7343
1. Direct evaluation of downstream layer reuse (W1)
We thank the reviewer for the constructive suggestion. We conducted a method-blind human evaluation involving 50 evaluators with professional design backgrounds, each reviewing the same 500 cases randomly sampled from Crello-Testset and ScaleEdit-Testset. Each case includes at least three successive rounds of layer reuse, with edits carried forward between rounds. Operations include insertion, deletion, transformation, reordering, and text-layer editing. All methods underwent the same sequences with the same parameters. The modified stacks are recomposed without regeneration or repair, allowing evaluators to inspect layer boundaries, residual fragments, and newly exposed content (Section 4.6, p. 10; Appendix C.4, p. 23).
As shown in Table 5 (p. 10), LayerEditor achieves 7.97 for editing reusability, compared with 6.13 without C-EPD and 4.81 for the strongest external baseline. We found that unintended content mixing and blurred layer boundaries commonly hinder competing methods during subsequent reuse. Figure 8 (p. 10) shows independent movement and resizing of the turkey and basket. Figures 15–16 (pp. 27–28) show different requested layer organizations and successive manipulation of the resulting layers.
Table 5. Human evaluation of layered editing (p. 10). Scores range from 1 to 10, with higher values indicating better results. Edit., VQ., Cont., Comp., and Reuse. denote edit consistency, visual quality, layer-content accuracy, layer-composition accuracy, and editing reusability, respectively.