# AlphaVAE: Unified End-to-End RGBA Image Reconstruction and Generation with Alpha-Aware Representation Learning

arXiv: https://arxiv.org/abs/2507.09308 · HTML: https://mlsanity.com/paper/2507.09308/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models.
- Builds on 27 of its 41 references that are in the map.
- First posted on arXiv 12 Jul 2025.

## Placed in

- Diffusion models / [Diffusion foundations and sampling](https://mlsanity.com/diffusion/diffusion-foundations-and-sampling/)

## Builds on (most cited first, 27 of 27)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Improved Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2102.09672/) (2021)
- [Taming Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2012.09841/) (2020)
- [Real-Time High-Resolution Background Matting](https://mlsanity.com/paper/2012.07810/) (2020)
- [Deep Image Matting](https://mlsanity.com/paper/1703.03872/) (2017)
- [Transparent Image Layer Diffusion using Latent Transparency](https://mlsanity.com/paper/2402.17113/) (2024)
- [Bridging Composite and Real: Towards End-to-end Deep Image Matting](https://mlsanity.com/paper/2010.16188/) (2020)
- [Mask Guided Matting via Progressive Refinement Network](https://mlsanity.com/paper/2012.06722/) (2020)
- [Privacy-Preserving Portrait Matting](https://mlsanity.com/paper/2104.14222/) (2021)
- [Semantic Image Matting](https://mlsanity.com/paper/2104.08201/) (2021)
- [Text2Layer: Layered Image Generation using Latent Diffusion Model](https://mlsanity.com/paper/2307.09781/) (2023)
- [Generative Image Layer Decomposition with Visual Effects](https://mlsanity.com/paper/2411.17864/) (2024)
- [Rethinking Portrait Matting with Privacy Preserving](https://mlsanity.com/paper/2203.16828/) (2022)
- [Referring Image Matting](https://mlsanity.com/paper/2206.05149/) (2022)
- [Alfie: Democratising RGBA Image Generation With No ](https://mlsanity.com/paper/2408.14826/) (2024)
- [ZIM: Zero-Shot Image Matting for Anything](https://mlsanity.com/paper/2411.00626/) (2024)
- [TransMatting: Enhancing Transparent Objects Matting with Transformers](https://mlsanity.com/paper/2208.03007/) (2022)
- [Generating Compositional Scenes via Text-to-image RGBA Instance Generation](https://mlsanity.com/paper/2411.10913/) (2024)
- [Fast Soft Color Segmentation](https://mlsanity.com/paper/2004.08096/) (2020)
- [Towards Natural Image Matting in the Wild via Real-Scenario Prior](https://mlsanity.com/paper/2410.06593/) (2024)

## Built on by (most cited first, 8 of 8)

- [Qwen-Image-Layered: Towards Inherent Editability via Layer Decomposition](https://mlsanity.com/paper/2512.15603/) (2025)
- [Wan-Image: Pushing the Boundaries of Generative Visual Intelligence](https://mlsanity.com/paper/2604.19858/) (2026)
- [OmniPSD: Layered PSD Generation with Diffusion Transformer](https://mlsanity.com/paper/2512.09247/) (2025)
- [Video Generation with Stable Transparency via Shiftable RGB-A Distribution Learner](https://mlsanity.com/paper/2509.24979/) (2025)
- [Does Synthetic Layered Design Data Benefit Layered Design Decomposition?](https://mlsanity.com/paper/2605.15167/) (2026)
- [LimeCross: Context-Conditioned Layered Image Editing with Structural Consistency](https://mlsanity.com/paper/2605.10319/) (2026)
- [UniWorld-Design: From Pixel Generation to Layer-Native Design](https://mlsanity.com/paper/2608.03971/) (2026)
- [PixelART: Image-to-Layer Decomposition without Latents or Text-to-Image Pretraining](https://mlsanity.com/paper/2609.22789/) (2026)

## Abstract (arXiv)

Recent advances in latent diffusion models have achieved remarkable results in high-fidelity RGB image synthesis by leveraging pretrained VAEs to compress and reconstruct pixel data at low computational cost. However, the generation of transparent or layered content (RGBA image) remains largely unexplored, due to the lack of large-scale benchmarks. In this work, we propose ALPHA, the first comprehensive RGBA benchmark that adapts standard RGB metrics to four-channel images via alpha blending over canonical backgrounds. We further introduce ALPHAVAE, a unified end-to-end RGBA VAE that extends a pretrained RGB VAE by incorporating a dedicated alpha channel. The model is trained with a composite objective that combines alpha-blended pixel reconstruction, patch-level fidelity, perceptual consistency, and dual KL divergence constraints to ensure latent fidelity across both RGB and alpha representations. Our RGBA VAE, trained on only 8K images in contrast to 1M used by prior methods, achieves a +4.9 dB improvement in PSNR and a +3.2% increase in SSIM over LayerDiffuse in reconstruction. It also enables superior transparent image generation when fine-tuned within a latent diffusion framework. Our code, data, and models are released on https://github.com/o0o0o00o0/AlphaVAE for reproducibility.