<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Tong開發日記 on Medium]]></title>
        <description><![CDATA[Stories by Tong開發日記 on Medium]]></description>
        <link>https://medium.com/@qazwsx04152?source=rss-2c40778919bf------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*t9QlS9K_XxMy998gK3l8sg.png</url>
            <title>Stories by Tong開發日記 on Medium</title>
            <link>https://medium.com/@qazwsx04152?source=rss-2c40778919bf------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 06 Jul 2026 17:58:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@qazwsx04152/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com" rel="hub"/>
        <item>
            <title><![CDATA[（新手向）ComfyUI/Tripo3D/TouchDesigner《Ai生成的模型自動傳到Touchdesigner?》]]></title>
            <link>https://medium.com/@qazwsx04152/%E5%B0%88%E6%A1%88%E9%96%8B%E7%99%BC%E7%A2%8E%E7%89%87%E4%B9%8B-touchdesigner%E5%A6%82%E4%BD%95%E8%B7%9Fcomfyui-tripo3d%E9%80%A3%E5%8B%95-%E4%BD%BF%E7%94%A8made-from-fruits-sdxl-by-sa-may-650471c3a09d?source=rss-2c40778919bf------2</link>
            <guid isPermaLink="false">https://medium.com/p/650471c3a09d</guid>
            <category><![CDATA[touchdesigner]]></category>
            <category><![CDATA[3d-modeling]]></category>
            <category><![CDATA[generative-ai-tools]]></category>
            <category><![CDATA[generative-art]]></category>
            <category><![CDATA[comfy-ui]]></category>
            <dc:creator><![CDATA[Tong開發日記]]></dc:creator>
            <pubDate>Wed, 02 Apr 2025 19:53:23 GMT</pubDate>
            <atom:updated>2025-04-03T20:09:14.621Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ub2hdCSAHLm2uXHM.jpeg" /></figure><h3>前言</h3><blockquote>初次見面，我是阿艟TongTong 。先做個自我介紹:目前是就讀北科互動設計系的大二生，上學期，也就是2024年的冬天，我初次接觸了TouchDesigner。它以節點式的可視化邏輯，吸引了我的注意，並讓我藉由軟體，與團隊完成了第一個完整的TD成人向專案《孔外狂徒》</blockquote><p>作品連結:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.behance.net%2Fembed%2Fproject%2F220907841%3Filo0%3D1&amp;display_name=Behance&amp;url=https%3A%2F%2Fwww.behance.net%2Fgallery%2F220907841%2F_%2Fmodules%2F1260115719&amp;image=https%3A%2F%2Fa5.behance.net%2Fd8845d3def14f8c5ae187699eae2bf1e7bb3acd7%2Fimg%2Fcovers%2F808-blocked.png&amp;type=text%2Fhtml&amp;schema=behance" width="404" height="316" frameborder="0" scrolling="no"><a href="https://medium.com/media/8beca42b9679eeb5cbe1c4a7a3243e95/href">https://medium.com/media/8beca42b9679eeb5cbe1c4a7a3243e95/href</a></iframe><p>在2025年的春天，我想更進一步，將TD與更多軟體進行串接~若是數位藝術與AI能夠結合在一起，那設計師的玩具又要多不少了(邪惡偷笑)</p><h3>這個workflow展示的是Ai參考我的臉去生成蔬果臉，再變成3D蔬果面具。</h3><h3>對，就是這麼簡單粗暴。</h3><p>總之，這只是大二專題的一部分，是技術的碎片。但因為各種原因，暫時不會繼續優化，我就先把基礎的workflow放上來。</p><blockquote>這個玩具的概念: AI捕捉我的臉，生成一個蔬果臉模型，顯示在TD，以便進行更多藝術加工。</blockquote><blockquote>目標流程: <strong>ComfyUI的AI產生圖</strong>&gt;<strong>ComfyUI轉存模型檔&gt;TD讀取/顯示檔案</strong></blockquote><h3>一、ComfyUI配置環境</h3><h4>1. ComfyUI本體:秋葉整合包0.3.27</h4><blockquote>※整合包網址，視需求更新到對應版本即可，這裡我先用0.3.27</blockquote><p><a href="https://www.bilibili.com/video/BV1Ew411776J/?spm_id_from=333.337.search-card.all.click&amp;vd_source=f618272e7faaf655f3cccaa967e4b0bd">【AI绘画】ComfyUI整合包发布！解压即用 一键启动 工作流版界面 超多节点 ☆更新 ☆汉化 秋叶整合包_哔哩哔哩_bilibili</a></p><blockquote><strong><em>ComfyUI manager中安裝客製插件:</em></strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3CyLZrP6i-i_822LNg6h7A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0x0p1zWK5SUtI331H59k-w.png" /></figure><h4>2–0. WorkFlow偷懶大法-直接從ComfyUI的範例Sample下去改</h4><blockquote>接下來跟著這位大佬一步一步做就好，不多做贅述。</blockquote><p><a href="https://www.bilibili.com/video/BV15Z421H7F8/?spm_id_from=333.337.search-card.all.click&amp;vd_source=f618272e7faaf655f3cccaa967e4b0bd">Triposr本地部署 AI绘图+AI建模 一套搞定~Triposr for SDcomfyUI~_哔哩哔哩_bilibili</a></p><h4>※※ 如果有報錯請再看我這邊※※</h4><blockquote>MixLab跟Triopo都是生成3D模型的插件，聽說ComfyUI更新後，<strong>有些TripoSR的插件會報錯，所以額外裝了MixLab來補。</strong></blockquote><p>當然…如果你很幸運的<strong>沒有報錯。那可以只裝TripoSR Node，再直接下載它的exsample就好</strong>，那更簡單。只是我不幸運嗚嗚…</p><p><strong>如果你有報錯的話，請根據下面的影片進行操作，或是上GitHub尋求解答。</strong></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FTSNOsCCv40A%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DTSNOsCCv40A&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FTSNOsCCv40A%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/6ed88ea914aa735502d3df96fb97951f/href">https://medium.com/media/6ed88ea914aa735502d3df96fb97951f/href</a></iframe><h4>2–1. ComfyUI 完整workflow(左)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rCMEafG4X7RjxDxGaj-qog.png" /><figcaption>由左至右跑workflow。<br>1是使用的lora風格模型<br>2是參考攝像頭(也就是我的臉)<br>3是Ai各種主要參數的調整</figcaption></figure><h4>2–2. ComfyUI 完整workflow-2(右)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-K-cu39bVxU_xI-pwE6J0Q.png" /><figcaption>5、6是移除2d圖片背景<br>7根據去背後的圖片生成3d模型面具<br>8可以預覽面具樣貌<br>9負責把圖片存到檔案夾中(每生成一次就會新增檔案)</figcaption></figure><h4>3–1下載網上的Lora蔬果臉風格模型</h4><blockquote>這個Lora model的Base Model是SDXL 1.0</blockquote><blockquote>請確保你有先從HuggingFace官網下載<strong>SDXL 1.0，</strong>拉入<strong>ComfyUI\models\checkpoints</strong>資料夾中</blockquote><p><a href="https://civitai.com/models/137107/made-from-fruits-sdxl">Made from fruits. SDXL - v1.0 | Stable Diffusion XL LoRA | Civitai</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CvK9Ks50iFxnrOS2-5XRlQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-r2NQOOIoEOM5FbvlF1LBw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3Y45q-jGnEfXh1ImPcs3lQ.png" /></figure><p>※Load Checkpoint那邊有了正確的<strong>Base Model</strong>，模型才能正確運行喔。</p><p>不然會報錯。</p><h4>4. Finally!! 按下「Queue Prompt」讓你的ai蔬果臉面具被創造吧!</h4><blockquote>我圈的節點如果有接到TripoSRSampler的話(如圖)</blockquote><blockquote>生成的蔬果面具‧obj就會傳到資料夾C:\Users\user\Desktop\coomfyUInew\ComfyUI\output</blockquote><p>※雖然這個節點不會顯示預覽，但上面接的Save TripoSR mesh是可以預覽的，影響不大(可能又是Bug吧...)。</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uFxxKeQDULMqCyFUPyEW-A.png" /></figure><p>生成的<strong>面具們.obj</strong>會被儲存在這邊</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nyHq2jujFhfZlrOySf94-w.png" /></figure><h3>TouchDesigner:</h3><h4><strong>1. 加一個folder DAT把路徑改成剛剛的Output資料夾，並連到一個select2 DAT node</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Z2PaepBf05zzqK6jiOqaJQ.png" /></figure><h4><strong>2. 修改select2 DAT node參數到跟我一樣，這個Node會選取檔案夾中最新的obj檔案(流水數字最大的檔案名稱)</strong></h4><p>Start Row Index:</p><pre>me.inputTable.numRows - 1</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QPaTMrh1PRl4WtF7YIbL7g.png" /></figure><h4>3.調整一下位置角度，現在你每次生成的面具，都能在TD上即時顯示~</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7OJ83Q2Q5aYunxRlU6B9xA.png" /></figure><p>今天的專題進度就到這樣，若後續有做優化又剛好有空更新，就會傳到這裡!</p><p>By-清明節很想睡覺但又要考多益還要找實習、做專題的悲催阿艟</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=650471c3a09d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>