{
  "@context": "https://schema.org",
  "@id": "https://aggr.aymericbeaumet.com/items/qwen-ai/2025-12-19-qwen-image-layered-layered-decomposition-for-inherent/#webpage",
  "@type": [
    "WebPage",
    "ArchiveComponent"
  ],
  "articleBody": "![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/layered.JPG)\n\n[QWEN CHAT](https://qwen.ai) [GITHUB](https://github.com/QwenLM/Qwen-Image-Layered) [HUGGING FACE](https://huggingface.co/Qwen/Qwen-Image-Layered) [MODELSCOPE](https://modelscope.cn/models/Qwen/Qwen-Image-Layered) [DEMO](https://https://huggingface.co/spaces/Qwen/Qwen-Image-Layered)\n\nToday, we are excited to introduce Qwen-Image-Layered, a model capable of decomposing an image into multiple RGBA layers. This layered representation unlocks inherent editability: each layer can be independently manipulated without affecting other content. Meanwhile, such a layered representation naturally supports high-fidelity elementary operations-such as resizing, reposition, and recoloring. By physically isolating semantic or structural components into distinct layers, our approach enables high-fidelity and consistent editing.\n\n## Layered Decomposition in Application\n\nGiven an image, Qwen-Image-Layered can decompose it into several RGBA layers:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%871.JPG#center%20)\n\nAfter decomposition, edits are applied exclusively to the target layer, physically isolating it from the rest of the content, and thereby fundamentally ensuring consistency across edits.\n\nFor example, we can recolor the first layer and keep all other content untouched:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%872.JPG#center%20)\n\nWe can also replace the second layer from a girl to a boy:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%873.JPG#center%20)\n\nHere, we revise the text to “Qwen-Image”:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%874.JPG#center%20)\n\nFurthermore, the layered structure naturally supports elemetary operations. For example, we can delete unwanted objects cleanly:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%875.JPG#center%20)\n\nWe can also resize an object without distortion:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%876.JPG#center%20)\n\nAfter layer decomposition, we can move objects freely within the canvas:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%877.JPG#center%20)\n\n## Flexible and Iterative Decomposition\n\nQwen-Image-Layered is not limited to a fixed number of layers. The model supports variable-layer decomposition. For example, we can decompose an image into either 3 or 8 layers as needed:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%878.JPG#center%20)\n\nMoreover, decomposition can be applied recursively: any layer can itself be further decomposed, enabling infinite decomposition.\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%879.JPG#center%20)\n\n## Conclusion\n\nQwen-Image-Layered bridges the gap between raster imagery and structured, editable representations. By reimagining images as composable layers, we hope to enable intuitive, precise, and robust editing capabilities.\n\n## Citation\n\nIf you find our model useful in your research, please consider citing us 📝 :)\n\n```bibtex\n@misc{yin2025qwenimagelayered,\n      title={Qwen-Image-Layered: Towards Inherent Editability via Layer Decomposition}, \n      author={Shengming Yin, Zekai Zhang, Zecheng Tang, Kaiyuan Gao, Xiao Xu, Kun Yan, Jiahao Li, Yilei Chen, Yuxiang Chen, Heung-Yeung Shum, Lionel M. Ni, Jingren Zhou, Junyang Lin, Chenfei Wu},\n      year={2025},\n      eprint={2512.15603},\n      archivePrefix={arXiv},\n      primaryClass={cs.CV},\n      url={https://arxiv.org/abs/2512.15603}, \n}\n```\n",
  "authors": [
    "QwenTeam"
  ],
  "captured_at": "2026-10-18T23:27:08.867657739+00:00",
  "category": "ai",
  "content_html": "<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/layered.JPG\" alt=\"\" loading=\"eager\" decoding=\"async\" fetchpriority=\"high\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p><a target=\"_blank\" rel=\"noopener noreferrer\" href=\"https://qwen.ai\">QWEN CHAT</a> <a target=\"_blank\" rel=\"noopener noreferrer\" href=\"https://github.com/QwenLM/Qwen-Image-Layered\">GITHUB</a> <a target=\"_blank\" rel=\"noopener noreferrer\" href=\"https://huggingface.co/Qwen/Qwen-Image-Layered\">HUGGING FACE</a> <a target=\"_blank\" rel=\"noopener noreferrer\" href=\"https://modelscope.cn/models/Qwen/Qwen-Image-Layered\">MODELSCOPE</a> <a target=\"_blank\" rel=\"noopener noreferrer\" href=\"https://https://huggingface.co/spaces/Qwen/Qwen-Image-Layered\">DEMO</a></p>\n<p>Today, we are excited to introduce Qwen-Image-Layered, a model capable of decomposing an image into multiple RGBA layers. This layered representation unlocks inherent editability: each layer can be independently manipulated without affecting other content. Meanwhile, such a layered representation naturally supports high-fidelity elementary operations-such as resizing, reposition, and recoloring. By physically isolating semantic or structural components into distinct layers, our approach enables high-fidelity and consistent editing.</p>\n<h2>Layered Decomposition in Application</h2>\n<p>Given an image, Qwen-Image-Layered can decompose it into several RGBA layers:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%871.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>After decomposition, edits are applied exclusively to the target layer, physically isolating it from the rest of the content, and thereby fundamentally ensuring consistency across edits.</p>\n<p>For example, we can recolor the first layer and keep all other content untouched:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%872.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>We can also replace the second layer from a girl to a boy:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%873.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>Here, we revise the text to “Qwen-Image”:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%874.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>Furthermore, the layered structure naturally supports elemetary operations. For example, we can delete unwanted objects cleanly:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%875.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>We can also resize an object without distortion:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%876.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>After layer decomposition, we can move objects freely within the canvas:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%877.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<h2>Flexible and Iterative Decomposition</h2>\n<p>Qwen-Image-Layered is not limited to a fixed number of layers. The model supports variable-layer decomposition. For example, we can decompose an image into either 3 or 8 layers as needed:</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%878.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<p>Moreover, decomposition can be applied recursively: any layer can itself be further decomposed, enabling infinite decomposition.</p>\n<p><picture class=\"article-picture article-picture-fallback\" style=\"--image-ratio:16 / 9\"><img src=\"https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%879.JPG#center%20\" alt=\"\" loading=\"lazy\" decoding=\"async\" fetchpriority=\"low\" referrerpolicy=\"no-referrer\" class=\"progressive-image\"></picture></p>\n<h2>Conclusion</h2>\n<p>Qwen-Image-Layered bridges the gap between raster imagery and structured, editable representations. By reimagining images as composable layers, we hope to enable intuitive, precise, and robust editing capabilities.</p>\n<h2>Citation</h2>\n<p>If you find our model useful in your research, please consider citing us 📝 :)</p>\n<pre><code class=\"language-bibtex\"><span class=\"code-snippet\" data-language=\"Bibtex\"><span class=\"syntax-text syntax-bibtex\"><span class=\"syntax-meta syntax-entry syntax-braces syntax-bibtex\"><span class=\"syntax-keyword syntax-other syntax-entry-type syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-keyword syntax-bibtex\">@</span>misc</span><span class=\"syntax-punctuation syntax-section syntax-entry syntax-begin syntax-bibtex\">{</span><span class=\"syntax-entity syntax-name syntax-type syntax-entry-key syntax-bibtex\">yin2025qwenimagelayered</span>,\n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">title</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>Qwen-Image-Layered: Towards Inherent Editability via Layer Decomposition<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>, \n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">author</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>Shengming Yin, Zekai Zhang, Zecheng Tang, Kaiyuan Gao, Xiao Xu, Kun Yan, Jiahao Li, Yilei Chen, Yuxiang Chen, Heung-Yeung Shum, Lionel M. Ni, Jingren Zhou, Junyang Lin, Chenfei Wu<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>,\n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">year</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>2025<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>,\n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">eprint</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>2512.15603<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>,\n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">archivePrefix</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>arXiv<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>,\n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">primaryClass</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>cs.CV<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>,\n      <span class=\"syntax-meta syntax-key-assignment syntax-bibtex\"><span class=\"syntax-string syntax-unquoted syntax-key syntax-bibtex\">url</span><span class=\"syntax-punctuation syntax-separator syntax-key-value syntax-bibtex\">=</span><span class=\"syntax-string syntax-quoted syntax-other syntax-braces syntax-bibtex\"><span class=\"syntax-punctuation syntax-definition syntax-string syntax-begin syntax-bibtex\">{</span>https://arxiv.org/abs/2512.15603<span class=\"syntax-punctuation syntax-definition syntax-string syntax-end syntax-bibtex\">}</span></span></span>, \n<span class=\"syntax-punctuation syntax-section syntax-entry syntax-end syntax-bibtex\">}</span></span>\n</span></span></code></pre>\n",
  "content_markdown": "![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/layered.JPG)\n\n[QWEN CHAT](https://qwen.ai) [GITHUB](https://github.com/QwenLM/Qwen-Image-Layered) [HUGGING FACE](https://huggingface.co/Qwen/Qwen-Image-Layered) [MODELSCOPE](https://modelscope.cn/models/Qwen/Qwen-Image-Layered) [DEMO](https://https://huggingface.co/spaces/Qwen/Qwen-Image-Layered)\n\nToday, we are excited to introduce Qwen-Image-Layered, a model capable of decomposing an image into multiple RGBA layers. This layered representation unlocks inherent editability: each layer can be independently manipulated without affecting other content. Meanwhile, such a layered representation naturally supports high-fidelity elementary operations-such as resizing, reposition, and recoloring. By physically isolating semantic or structural components into distinct layers, our approach enables high-fidelity and consistent editing.\n\n## Layered Decomposition in Application\n\nGiven an image, Qwen-Image-Layered can decompose it into several RGBA layers:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%871.JPG#center%20)\n\nAfter decomposition, edits are applied exclusively to the target layer, physically isolating it from the rest of the content, and thereby fundamentally ensuring consistency across edits.\n\nFor example, we can recolor the first layer and keep all other content untouched:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%872.JPG#center%20)\n\nWe can also replace the second layer from a girl to a boy:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%873.JPG#center%20)\n\nHere, we revise the text to “Qwen-Image”:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%874.JPG#center%20)\n\nFurthermore, the layered structure naturally supports elemetary operations. For example, we can delete unwanted objects cleanly:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%875.JPG#center%20)\n\nWe can also resize an object without distortion:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%876.JPG#center%20)\n\nAfter layer decomposition, we can move objects freely within the canvas:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%877.JPG#center%20)\n\n## Flexible and Iterative Decomposition\n\nQwen-Image-Layered is not limited to a fixed number of layers. The model supports variable-layer decomposition. For example, we can decompose an image into either 3 or 8 layers as needed:\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%878.JPG#center%20)\n\nMoreover, decomposition can be applied recursively: any layer can itself be further decomposed, enabling infinite decomposition.\n\n![](https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/layered/%E5%B9%BB%E7%81%AF%E7%89%879.JPG#center%20)\n\n## Conclusion\n\nQwen-Image-Layered bridges the gap between raster imagery and structured, editable representations. By reimagining images as composable layers, we hope to enable intuitive, precise, and robust editing capabilities.\n\n## Citation\n\nIf you find our model useful in your research, please consider citing us 📝 :)\n\n```bibtex\n@misc{yin2025qwenimagelayered,\n      title={Qwen-Image-Layered: Towards Inherent Editability via Layer Decomposition}, \n      author={Shengming Yin, Zekai Zhang, Zecheng Tang, Kaiyuan Gao, Xiao Xu, Kun Yan, Jiahao Li, Yilei Chen, Yuxiang Chen, Heung-Yeung Shum, Lionel M. Ni, Jingren Zhou, Junyang Lin, Chenfei Wu},\n      year={2025},\n      eprint={2512.15603},\n      archivePrefix={arXiv},\n      primaryClass={cs.CV},\n      url={https://arxiv.org/abs/2512.15603}, \n}\n```\n",
  "dateCreated": "2026-10-18T23:27:08.867657739+00:00",
  "date_created": "2025-12-19T05:08:30+00:00",
  "date_modified": null,
  "date_published": "2025-12-19T05:08:30+00:00",
  "description": "QWEN CHAT GITHUB HUGGING FACE MODELSCOPE DEMO Today, we are excited to introduce Qwen-Image-Layered, a model capable of decomposing an image into multiple RGBA layers. This layered representation unlocks inherent editability: each layer…",
  "external_url": "https://qwen.ai/blog?id=qwen-image-layered",
  "git": {
    "commit": "ebb0941c5c06a2ab2377342d607535262a3e969e",
    "path": "items/qwen-ai/2025/12/2025-12-19-qwen-image-layered-layered-decomposition-for-inherent.md",
    "permalink": "https://github.com/aymericbeaumet/aggr-instance/blob/ebb0941c5c06a2ab2377342d607535262a3e969e/items/qwen-ai/2025/12/2025-12-19-qwen-image-layered-layered-decomposition-for-inherent.md"
  },
  "id": "urn:aggr:item:56611e32fbd6c56d17b2c791da86096352a578aa",
  "isBasedOn": {
    "@id": "https://qwen.ai/blog?id=qwen-image-layered"
  },
  "mainEntity": {
    "@id": "https://qwen.ai/blog?id=qwen-image-layered",
    "@type": "CreativeWork",
    "archivedAt": {
      "@id": "https://aggr.aymericbeaumet.com/items/qwen-ai/2025-12-19-qwen-image-layered-layered-decomposition-for-inherent/#webpage"
    },
    "author": [
      {
        "@type": "Person",
        "name": "QwenTeam"
      }
    ],
    "datePublished": "2025-12-19T05:08:30+00:00",
    "headline": "Qwen-Image-Layered: Layered Decomposition for Inherent Editablity",
    "image": {
      "@type": "ImageObject",
      "caption": null,
      "height": 90,
      "url": "https://img.alicdn.com/imgextra/i2/O1CN01y9LKxE248WbyrZzFw_!!6000000007346-0-tps-1590-954.jpg",
      "width": 160
    },
    "timeRequired": "PT2M",
    "url": "https://qwen.ai/blog?id=qwen-image-layered",
    "wordCount": 344
  },
  "name": "Archived snapshot: Qwen-Image-Layered: Layered Decomposition for Inherent Editablity",
  "preview": {
    "alt": null,
    "color": null,
    "height": 90,
    "placeholder": {
      "data_url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mNk+A8AAQUBAScY42YAAAAASUVORK5CYII=",
      "hash": ""
    },
    "url": "https://img.alicdn.com/imgextra/i2/O1CN01y9LKxE248WbyrZzFw_!!6000000007346-0-tps-1590-954.jpg",
    "width": 160
  },
  "reading_minutes": 2,
  "replicated_at": null,
  "source": "qwen-ai",
  "source_name": "Qwen",
  "summary": null,
  "tags": [
    "research"
  ],
  "temporalCoverage": "2026-10-18T23:27:08.867657739+00:00",
  "title": "Qwen-Image-Layered: Layered Decomposition for Inherent Editablity",
  "url": "https://aggr.aymericbeaumet.com/items/qwen-ai/2025-12-19-qwen-image-layered-layered-decomposition-for-inherent/",
  "word_count": 344
}