Buckets:

hf-doc-build
/

doc

Files
xet
hf-doc-build/doc / diffusers /main /ja /advanced_inference /outpaint.html
mishig
3 days ago
download
raw
33.5 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;アウトペインティング&quot;,&quot;local&quot;:&quot;outpainting&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;画像の準備&quot;,&quot;local&quot;:&quot;image-preparation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Outpaint&quot;,&quot;local&quot;:&quot;outpaint&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
		<link href="/docs/diffusers/main/en/_app/immutable/entry/start.0S0_iEoG.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/I9eooN77.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/Bupkjr9f.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/entry/app.CPdOxCbw.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/CyvZ3FS5.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/l0QnnOED.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DnVIJfC5.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/0.DteTvmQZ.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/CoSrdQax.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/2.CJ7qCPpK.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DSPDl1gz.js" rel="modulepreload">
		<!--w57hbo--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Outpainting&quot;,&quot;local&quot;:&quot;outpainting&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Image preparation&quot;,&quot;local&quot;:&quot;image-preparation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Outpaint&quot;,&quot;local&quot;:&quot;outpaint&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/><!---->
		<link href="/docs/diffusers/main/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: inline-end; margin-inline-start: 10px; display: inline-flex; position: relative; z-index: 10;"><div dir="ltr" class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----></span> <span dir="auto">Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg><!----></button></div> <!--[-1--><!--]--></div><!----> <!--[0--><h1 class="relative group"><a id="outpainting" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#outpainting"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->アウトペインティング<!----></span></h1><!--]--><!----> <p>Outpaintingは、画像を元の境界を超えて拡張する技術であり、元の画像を保持したまま、画像内の視覚要素を追加、置換、または変更できます。<a href="../using-diffusers/inpaint">Inpainting</a>と同様に、元の画像（黒ピクセルのマスクで表されます）を維持しながら、白い領域（この場合は元の画像の外側）を新しい視覚要素で埋めたいです。Outpaintingにはいくつかの方法があり、<a href="https://hf.co/blog/OzzyGT/outpainting-controlnet" rel="nofollow">ControlNet</a>や<a href="https://hf.co/blog/OzzyGT/outpainting-differential-diffusion" rel="nofollow">Differential Diffusion</a>などがあります。</p> <p>このガイドでは、インペインティングモデル、ControlNet、およびZoeDepth推定器を使用してアウトペインティングを行う方法を紹介します。</p> <p>開始する前に、ZoeDepth 推定器を使用できるように、<a href="https://github.com/huggingface/controlnet_aux" rel="nofollow">controlnet_aux</a> ライブラリがインストールされていることを確認してください。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->!pip install -q controlnet_aux<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="image-preparation" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#image-preparation"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->画像の準備<!----></span></h2><!--]--><!----> <p>アウトペイントに使用する画像を選択し、<a href="https://hf.co/spaces/briaai/BRIA-RMBG-1.4" rel="nofollow">BRIA-RMBG-1.4</a>のようなSpaceで背景を除去することから始めます。</p> <iframe src="https://briaai-bria-rmbg-1-4.hf.space" frameborder="0" width="850" height="450"></iframe> <p>例えば、この靴のペアの画像から背景を削除します。</p> <div class="flex flex-row gap-4"><div class="flex-1"><img class="rounded-xl" src="https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/original-jordan.png"/> <figcaption class="mt-2 text-center text-sm text-gray-500">元の画像</figcaption></div> <div class="flex-1"><img class="rounded-xl" src="https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/no-background-jordan.png"/> <figcaption class="mt-2 text-center text-sm text-gray-500">背景を削除しました</figcaption></div></div> <p><a href="../api/pipelines/stable_diffusion/stable_diffusion_xl">Stable Diffusion XL (SDXL)</a> モデルは 1024x1024 の画像で最も良い結果を得られますが、ハードウェアのメモリが十分であれば、画像を任意のサイズにリサイズできます。画像の透明な背景は、白い背景に置き換える必要があります。以下のような関数を作成し、画像をスケーリングして白い背景に貼り付けてください。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> random

<span class="hljs-keyword">import</span> requests
<span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> controlnet_aux <span class="hljs-keyword">import</span> ZoeDetector
<span class="hljs-keyword">from</span> PIL <span class="hljs-keyword">import</span> Image, ImageOps

<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> (
    AutoencoderKL,
    ControlNetModel,
    StableDiffusionXLControlNetPipeline,
    StableDiffusionXLInpaintPipeline,
)

<span class="hljs-keyword">def</span> <span class="hljs-title function_">scale_and_paste</span>(<span class="hljs-params">original_image</span>):
    aspect_ratio = original_image.width / original_image.height

    <span class="hljs-keyword">if</span> original_image.width &gt; original_image.height:
        new_width = <span class="hljs-number">1024</span>
        new_height = <span class="hljs-built_in">round</span>(new_width / aspect_ratio)
    <span class="hljs-keyword">else</span>:
        new_height = <span class="hljs-number">1024</span>
        new_width = <span class="hljs-built_in">round</span>(new_height * aspect_ratio)

    resized_original = original_image.resize((new_width, new_height), Image.LANCZOS)
    white_background = Image.new(<span class="hljs-string">&quot;RGBA&quot;</span>, (<span class="hljs-number">1024</span>, <span class="hljs-number">1024</span>), <span class="hljs-string">&quot;white&quot;</span>)
    x = (<span class="hljs-number">1024</span> - new_width) // <span class="hljs-number">2</span>
    y = (<span class="hljs-number">1024</span> - new_height) // <span class="hljs-number">2</span>
    white_background.paste(resized_original, (x, y), resized_original)

    <span class="hljs-keyword">return</span> resized_original, white_background

original_image = Image.<span class="hljs-built_in">open</span>(
    requests.get(
        <span class="hljs-string">&quot;https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/no-background-jordan.png&quot;</span>,
        stream=<span class="hljs-literal">True</span>,
    ).raw
).convert(<span class="hljs-string">&quot;RGBA&quot;</span>)
resized_img, white_bg_image = scale_and_paste(original_image)<!----></pre></div><!----> <p>不要な余分な詳細を追加することを避けるために、ZoeDepth推定器を使用して、生成中に追加のガイダンスを提供し、靴が元の画像と一貫性を保つようにします。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->zoe = ZoeDetector.from_pretrained(<span class="hljs-string">&quot;lllyasviel/Annotators&quot;</span>)
image_zoe = zoe(white_bg_image, detect_resolution=<span class="hljs-number">512</span>, image_resolution=<span class="hljs-number">1024</span>)
image_zoe<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/zoedepth-jordan.png"/></div> <!--[1--><h2 class="relative group"><a id="outpaint" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#outpaint"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->Outpaint<!----></span></h2><!--]--><!----> <p>画像の準備が整ったら、インペインティング用に学習された SDXL ControlNet である <a href="https://hf.co/destitech/controlnet-inpaint-dreamer-sdxl" rel="nofollow">controlnet-inpaint-dreamer-sdxl</a> を使用して、靴の周囲の白い領域にコンテンツを生成できます。</p> <p>インペインティング用の ControlNet、ZoeDepth モデル、VAE を読み込み、<a href="/docs/diffusers/main/ja/api/pipelines/controlnet_sdxl#diffusers.StableDiffusionXLControlNetPipeline">StableDiffusionXLControlNetPipeline</a> に渡します。その後、初期画像をアウトペイントするための任意の <code>generate_image</code> 関数（利便性のため）を作成できます。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->controlnets = [
    ControlNetModel.from_pretrained(
        <span class="hljs-string">&quot;destitech/controlnet-inpaint-dreamer-sdxl&quot;</span>, dtype=torch.float16, variant=<span class="hljs-string">&quot;fp16&quot;</span>
    ),
    ControlNetModel.from_pretrained(
        <span class="hljs-string">&quot;diffusers/controlnet-zoe-depth-sdxl-1.0&quot;</span>, dtype=torch.float16
    ),
]
vae = AutoencoderKL.from_pretrained(<span class="hljs-string">&quot;madebyollin/sdxl-vae-fp16-fix&quot;</span>, dtype=torch.float16).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline = StableDiffusionXLControlNetPipeline.from_pretrained(
    <span class="hljs-string">&quot;SG161222/RealVisXL_V4.0&quot;</span>, dtype=torch.float16, variant=<span class="hljs-string">&quot;fp16&quot;</span>, controlnet=controlnets, vae=vae
).to(<span class="hljs-string">&quot;cuda&quot;</span>)

<span class="hljs-keyword">def</span> <span class="hljs-title function_">generate_image</span>(<span class="hljs-params">prompt, negative_prompt, inpaint_image, zoe_image, seed: <span class="hljs-built_in">int</span> = <span class="hljs-literal">None</span></span>):
    <span class="hljs-keyword">if</span> seed <span class="hljs-keyword">is</span> <span class="hljs-literal">None</span>:
        seed = random.randint(<span class="hljs-number">0</span>, <span class="hljs-number">2</span>**<span class="hljs-number">32</span> - <span class="hljs-number">1</span>)

    generator = torch.Generator(device=<span class="hljs-string">&quot;cpu&quot;</span>).manual_seed(seed)

    image = pipeline(
        prompt,
        negative_prompt=negative_prompt,
        image=[inpaint_image, zoe_image],
        guidance_scale=<span class="hljs-number">6.5</span>,
        num_inference_steps=<span class="hljs-number">25</span>,
        generator=generator,
        controlnet_conditioning_scale=[<span class="hljs-number">0.5</span>, <span class="hljs-number">0.8</span>],
        control_guidance_end=[<span class="hljs-number">0.9</span>, <span class="hljs-number">0.6</span>],
    ).images[<span class="hljs-number">0</span>]

    <span class="hljs-keyword">return</span> image

prompt = <span class="hljs-string">&quot;nike air jordans on a basketball court&quot;</span>
negative_prompt = <span class="hljs-string">&quot;&quot;</span>

temp_image = generate_image(prompt, negative_prompt, white_bg_image, image_zoe, <span class="hljs-number">908097</span>)<!----></pre></div><!----> <p>元の画像を、初期のアウトペイント済み画像の上に貼り付けます。アウトペイントされた背景は、後続のステップで改善します。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->x = (<span class="hljs-number">1024</span> - resized_img.width) // <span class="hljs-number">2</span>
y = (<span class="hljs-number">1024</span> - resized_img.height) // <span class="hljs-number">2</span>
temp_image.paste(resized_img, (x, y), resized_img)
temp_image<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/initial-outpaint.png"/></div> <blockquote class="tip"><p>メモリが不足している場合は、ここでメモリを解放する良いタイミングです！</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->pipeline=<span class="hljs-literal">None</span>
torch.cuda.empty_cache()<!----></pre></div><!----></blockquote> <p>初期のアウトペイント画像が準備できたので、<a href="/docs/diffusers/main/ja/api/pipelines/stable_diffusion/stable_diffusion_xl#diffusers.StableDiffusionXLInpaintPipeline">StableDiffusionXLInpaintPipeline</a>を<a href="https://hf.co/SG161222/RealVisXL_V4.0" rel="nofollow">RealVisXL</a>モデルで読み込み、より高品質な最終アウトペイント画像を生成します。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->pipeline = StableDiffusionXLInpaintPipeline.from_pretrained(
    <span class="hljs-string">&quot;OzzyGT/RealVisXL_V4.0_inpainting&quot;</span>,
    dtype=torch.float16,
    variant=<span class="hljs-string">&quot;fp16&quot;</span>,
    vae=vae,
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span><!----></pre></div><!----> <p>最終的なアウトペイント画像用のマスクを準備します。元の画像とアウトペイントされた背景の間により自然な遷移を作成するために、マスクをぼかし、より滑らかにブレンドできるようにします。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!---->mask = Image.new(<span class="hljs-string">&quot;L&quot;</span>, temp_image.size)
mask.paste(resized_img.split()[<span class="hljs-number">3</span>], (x, y))
mask = ImageOps.invert(mask)
final_mask = mask.point(<span class="hljs-keyword">lambda</span> p: p &gt; <span class="hljs-number">128</span> <span class="hljs-keyword">and</span> <span class="hljs-number">255</span>)
mask_blurred = pipeline.mask_processor.blur(final_mask, blur_factor=<span class="hljs-number">20</span>)
mask_blurred<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/blurred-mask.png"/></div> <p>より良いプロンプトを作成し、<code>generate_outpaint</code> 関数に渡して、最終的なアウトペイント画像を生成します。同様に、元の画像を最終的なアウトペイント背景の上に貼り付けます。</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5   text-gray-600  " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> コピーしました</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">def</span> <span class="hljs-title function_">generate_outpaint</span>(<span class="hljs-params">prompt, negative_prompt, image, mask, seed: <span class="hljs-built_in">int</span> = <span class="hljs-literal">None</span></span>):
    <span class="hljs-keyword">if</span> seed <span class="hljs-keyword">is</span> <span class="hljs-literal">None</span>:
        seed = random.randint(<span class="hljs-number">0</span>, <span class="hljs-number">2</span>**<span class="hljs-number">32</span> - <span class="hljs-number">1</span>)

    generator = torch.Generator(device=<span class="hljs-string">&quot;cpu&quot;</span>).manual_seed(seed)

    image = pipeline(
        prompt,
        negative_prompt=negative_prompt,
        image=image,
        mask_image=mask,
        guidance_scale=<span class="hljs-number">10.0</span>,
        strength=<span class="hljs-number">0.8</span>,
        num_inference_steps=<span class="hljs-number">30</span>,
        generator=generator,
    ).images[<span class="hljs-number">0</span>]

    <span class="hljs-keyword">return</span> image

prompt = <span class="hljs-string">&quot;high quality photo of nike air jordans on a basketball court, highly detailed&quot;</span>
negative_prompt = <span class="hljs-string">&quot;&quot;</span>

final_image = generate_outpaint(prompt, negative_prompt, temp_image, mask_blurred, <span class="hljs-number">7688778</span>)
x = (<span class="hljs-number">1024</span> - resized_img.width) // <span class="hljs-number">2</span>
y = (<span class="hljs-number">1024</span> - resized_img.height) // <span class="hljs-number">2</span>
final_image.paste(resized_img, (x, y), resized_img)
final_image<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/stevhliu/testing-images/resolve/main/final-outpaint.png"/></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/diffusers/blob/main/docs/source/en/advanced_inference/outpaint.md" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg><!----> <span><span class="underline">Update</span> on GitHub</span></a><!----> <p></p><!--]--><!----><!--]--><!--]--><!--]--> <!--[-1--><!--]--><!--]-->
			
			<script>
				{
					__sveltekit_1qsdbwf = {
						base: "/docs/diffusers/main/en",
						assets: "/docs/diffusers/main/en"
					};
					const element = document.currentScript.parentElement;
					Promise.all([
						import("/docs/diffusers/main/en/_app/immutable/entry/start.0S0_iEoG.js"),
						import("/docs/diffusers/main/en/_app/immutable/entry/app.CPdOxCbw.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 2],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		
Xet Storage Details

Size:: 33.5 kB
Xet hash:: ec3a440b2b734ccc9e852a8b0728d396b6cedb546387ccdf9470f45a8079b208
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.