PixelART: Image-to-Layer Decomposition without Latents or Text-to-Image Pretraining

Zelin Jia; Zhao Zhang; Zhicong Tang; Yuhui Yuan; Shixia Liu

cs.CVSep 19, 2026

PixelART: Image-to-Layer Decomposition without Latents or Text-to-Image Pretraining

Authors: Zelin Jia, Zhao Zhang, Zhicong Tang, Yuhui Yuan, Shixia Liu

Organizations: School of Software, BNRist, Tsinghua University · Canva Research

Abstract

Image-to-layer decomposition converts a flattened image into editable RGBA layers, enabling element-level editing in design workflows. Existing diffusion-based systems typically adapt large pretrained text-to-image (T2I) models and introduce RGBA autoencoders or variable-layer architectural modules. We revisit this design choice and ask whether layer decomposition truly requires these heavyweight components. We introduce PixelART, a pixel-space rectified-flow Transformer trained from scratch for image-to-layer (I2L) decomposition. PixelART directly denoises regional RGBA pixel patches using a single-stream multi-modal diffusion Transformer, avoiding RGBA-VAEs, pretrained T2I backbones, and layer-specific decoders. We identify a key property of the task: high-noise timesteps determine layer assignment and coarse layer organization, while low-noise timesteps mainly refine color, alpha, texture, and boundaries. Based on this observation, we propose a terminal-boosted timestep sampling strategy to increase training coverage in the high-noise layer assignment regime. Trained on 4M multi-layer design templates, PixelART achieves state-of-the-art layer decomposition and composite reconstruction on Design-Multi-Layer-Bench and LICA with over 80% fewer parameters, 98% lower latency, and 85% lower memory than the recent Qwen-Image-Layered model. Ablation experiments show that pixel-space $\mathbf{x}$ -prediction, terminal-boosted timestep sampling, and data/model scaling are critical, while T2I pretraining brings marginal benefits to the I2L task.

Figures & tables

Figure 1: Comparison with previous Image-to-Layer decomposition systems. Left: ART [ 47 ] and its variant; Middle: Qwen-Image-Layered [ 60 ] ; Right: Ours. Our PixelART is much simpler and trained from scratch.

Figure 2: Image-to-layer decomposition results with our PixelART-H/32.

Figure 3: High-noise timesteps are critical for layer assignment. We compare layer states along the sampling trajectory for checkpoints with and without terminal boosting. The boosted model (1st row) assigns regions correctly at the earliest high-noise stage, whereas the unboosted model (2nd row) makes a persistent layer-assignment error. Later steps mainly refine RGBA details. Green/red boxes mark correct/incorrect assignments.

Figure 4: Terminal-boosted timestep distributions. We mix the base schedule with either a point mass at

t=0

or a uniform distribution

\mathcal{U}(0,\tau)

to increase the mass in the high-noise region.

Figure 5: User study. Comparison with previous SoTA methods and the commercial system Magic Layers [ 4 ] .

				Design-Multi-Layer-Bench			LICA
				Pixel PSNR $\uparrow$	VLM (GPT/Gem.) $\uparrow$		Pixel PSNR $\uparrow$	VLM (GPT/Gem.) $\uparrow$		Inference Efficiency
Method	T2I Pre.	VAE	#Params (Layout)	Comp.	Alpha	Visual	Comp.	Alpha	Visual	Time (s) $\downarrow$	Mem (GB) $\downarrow$
CLD [ 36 ]	✓	✓	14B (12+2)	29.79	4.24/4.83	4.50/4.68	28.84	4.21/4.78	4.38/4.62	58.7 (+0.7)	55.6 (+4.2)
Qwen-Image-Layered [ 60 ]	✓	✓	20B	32.46	4.30/4.73	4.14/4.37	31.99	4.22/4.65	3.93/4.20	213.4	72.4
PixelART	✗	✗	3.05B (1.05+2)	36.92	4.32 / 4.91	4.57 / 4.81	36.39	4.26 / 4.87	4.44 / 4.75	3.1 (+0.7)	6.6 (+4.2)
LayerD † [ 54 ]	✓	✓	0.27B	36.07	4.12/3.94	3.49/3.13	37.22	4.06/4.01	3.64/3.37	15.0	3.4

Table 1: System-level comparison . T2I Pre. : whether initialized from pretrained T2I. Pixel PSNR : PSNR between the alpha-composited reconstruction and the input image. VLM : scores (1–5) from GPT-5.5 / Gemini-3.1 Pro (separated by ‘/’) on alpha and visual quality. Efficiency : per-image latency and peak memory on one H200. Parenthesized values for PixelART and CLD report the additional layout detector overhead (#parameters, latency, GPU memory). LayerD † is non-diffusion (segmentation + inpainting), which inflates Composite PSNR but yields the lowest VLM Visual score. Bold = best within the diffusion-based group.

Figure 6: Qualitative comparison with state-of-the-art approaches. QIL: Qwen-Image-Layered.

Table 8

Panel of Figure 7 — Figure 7: Ablation visualizations. Left: qualitative effect of perceptual losses on background reconstruction. Right top: training loss curves of Muon vs. AdamW under the same recipe. Right bottom: I2L fine-tuning loss curves for T2I-pretrained vs. from-scratch initializations on PixelART-B.

Appendix figures & tables6 assets

Supplementary material from the paper’s appendix.

Appendix

	PixelART-B	PixelART-L	PixelART-H	PixelART (main)
Architecture
# Transformer blocks	12	24	32	32
hidden dim $d$	768	1024	1280	1280
# attention heads	12	16	16	16
patch size $P$	16			32
input noise scale	1			2

Appendix

Table 10: Configurations of PixelART variants. PixelART-B is the default for all ablations and is also the small model in the model-size ablation, alongside PixelART-L and PixelART-H. PixelART (main model) is the separately trained system-level model reported in Table 1 ; it shares the H scale but uses a coarser patch size and a three-stage resolution curriculum. Settings that are common across multiple columns are merged. The pixel-space versus latent-space comparison (Table 2 ) uses the same PixelART-B architecture, but the latter operates on the RGBA-VAE latents of Qwen-Image-Layered with a patch size of 2, with the parameter count differing by

\sim

1M due to the smaller patch size.

Figure 8: Web interface used in the user study. Each question shows the original input image, two anonymized layer-decomposition outputs (A/B), the two-dimension ternary choice, and optional fields for failure tags and free-text notes.

	Layer metrics			Composite metrics
Prediction	L1 $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	L1 $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$
$\mathbf{x}$ -pred	0.0453	32.88	0.8995	0.0086	38.55	0.9727
$\mathbf{v}$ -pred	0.0526	30.17	0.8691	0.0114	36.08	0.9441

Appendix

Table 11:

\mathbf{x}

- vs.

\mathbf{v}

-prediction at PixelART-H scale (

d{=}1280

, no bottleneck). Both runs use identical recipes apart from the prediction target. At this width,

\mathbf{v}

-prediction does converge but still trails

\mathbf{x}

-prediction, indicating that the manifold mismatch is not solely a capacity issue.

	Composite metrics			Layer metrics
Decoder	L1 $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	L1 $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$
Qwen-Image-Layered (official)	0.0132	32.52	0.9568	0.0363	31.66	0.9289
Ours (300M ViT)	0.0083	33.69	0.9686	0.0254	37.49	0.9569

Appendix

Table 12: RGBA-VAE decoder reconstruction quality on Design-Multi-Layer-Bench at

256\times 256

. Each decoder takes the official Qwen-Image-Layered encoder’s latent of a ground-truth RGBA layer and produces a reconstruction, which is compared against the original. Both decoders share the same encoder. Layer L1/PSNR are alpha-pre-multiplied (Appendix D ); layer SSIM is alpha-weighted.

	single	two	counting	colors	color	position	overall
	obj.	obj.			attr.
PixelART-B (BLIP-3o, T2I)	74.38	24.24	18.12	61.70	12.75	7.50	33.12

Appendix

Table 13: GenEval scores of the BLIP-3o-pretrained PixelART-B (168M params, batch 1024, 250K iters), evaluated before I2L fine-tuning. Overall is the per-task average. The model does not adopt any specialized optimization recipes common in dedicated T2I models, yet still produces non-degenerate outputs across all six task categories at a level consistent with this parameter scale and training budget. This checkpoint serves as the T2I-initialized starting point for the I2L ablation in Table 3 .

Figure 9: Zero-shot decomposition on natural images. Although trained only on multi-layer design templates, PixelART produces coherent foreground/background separation and plausible occluded-region completion on selected natural-image inputs.

Pixel-Space Diffusion Models

Pixel-space diffusion models generate or restore images by learning to denoise raw pixels rather than compressed latent codes.

See all 55 papers

UniWorld-Design: From Pixel Generation to Layer-Native Design
Aug 4, 2026Zongjian Li, Zhiyuan Yan, Chenxu Bai +9Image Generation Compositional Image Generation
UniWorld Team1,2∗
PixelDiT2: Representation-Grounded Pixel Diffusion Transformers
Sep 21, 2026Yongsheng Yu, Wei Xiong, Yichen Sheng +2Diffusion Model Guidance Diffusion Transformer
NVIDIA · University of Rochester
Editing Images into Reusable Layers via Compositional Edit-Prior Distillation
Oct 17, 2026Wen Xue, Yong Liu, Xueheng Li +5Image Editing Text-Guided Image Editing
South China University of Technology · Tsinghua University · University of Science and Technology of China +1

Aug 4, 2026J/K move · Enter open · S save

UniWorld-Design: From Pixel Generation to Layer-Native Design

Zongjian Li, Zhiyuan Yan, Chenxu Bai, Chen Chen+8

UniWorld Team1,2∗

Image Generation Compositional Image Generation Controllable Image Generation

Open paper AskPDF arXiv

We introduce UniWorld-Design, a framework that redefines image generation from flat pixel synthesis to structured visual composition, with semantic RGBA layers as the atomic units of generation, understanding, and editing. Our key insight is that pixels define how an image is rendered, whereas layers define how an image is created, understood, and edited. Just as human designers create and manipulate visual content through layers rather than raw pixels, UniWorld-Design equips multimodal generative models with a layer-native design space. UniWorld-Design comprises two models. The Text-to-RGBA (T2RGBA) model generates standalone RGBA assets directly from text. The Image-to-Layer (I2L) model conditions on a finished image, a global instruction and per-layer prompts, and jointly produces ordered, complete semantic RGBA layers. Its instruction interface supports top-level decomposition, recursive decomposition and targeted extraction, making layering an instruction-addressable operation for agentic editing. Because I2L learns complete semantic objects rather than visible-pixel partitions, its layers stay usable when moved or removed. On the Crello benchmark, I2L reduces per-layer RGB L1 error by 37% and achieves a 34% relative improvement in Alpha Soft IoU over Qwen-Image-Layered. Separately, T2RGBA achieves the highest CLIP Score, outperforming LayerDiffuse and OmniAlpha.

Abstract

Pixel-Space Diffusion Models

Explore similar work

UniWorld-Design: From Pixel Generation to Layer-Native Design

PixelDiT2: Representation-Grounded Pixel Diffusion Transformers

Editing Images into Reusable Layers via Compositional Edit-Prior Distillation