ComfyUI Community Gathers in Shibuya for First-Ever Meetup: Shaping the Future of Image Generation AI
ComfyUI, the node-based image generation framework making waves in the AI art scene, saw its development team, Comfy Org, host its first-ever meetup in Shibuya, Tokyo on September 27, 2024.
The event drew a crowd of 37 enthusiastic attendees, including ComfyUI users, developers, and AI art enthusiasts. The night was filled with excitement as they delved into the latest updates, future prospects, and innovative applications of ComfyUI.
Presentations by the Development Team: Unveiling New Features, Models, and the Future of ComfyUI
The evening commenced with an opening address by Yoland, the leader of Comfy Org, who warmly introduced the team and shared their passion for ComfyUI.
Next, HCL and Robin took the stage to unveil upcoming features for the beta version of Electron, the frontend, and Electron itself, generating a wave of anticipation and excitement among the attendees.
ComfyAnoymous then presented insights into new models, followed by kosinkadink, who shed light on advancements in ADE and model patching. These presentations provided a valuable glimpse into the ongoing evolution of ComfyUI.
Engaging Q&A Session: Deep Dive into ComfyUI and the Future of AI Image Generation
Following the presentations, attendees had the opportunity to engage in a Q&A session with the development team. The discussion was lively, covering topics ranging from practical aspects of using ComfyUI to broader questions about the future of AI image generation.
Lightning Talks & Workflow Sharing: Fostering Connections and Collaboration
The latter half of the event was dedicated to lightning talks (LTs) and workflow sharing by ComfyUI users.
The face seals were applied by the editor-in-chief of AICU media (human processed).
With a diverse range of topics related to ComfyUI, attendees showcased their creative applications and unique workflows. This segment fostered a valuable learning environment where participants shared knowledge, experiences, and inspiration.
★ Call for Contributions: We at AICU media encourage speakers who shared their insights at the meetup to consider contributing their presentations to our platform. We offer compensation for published articles.
Networking and After-Party: Building the ComfyUI Community
The evening concluded with a networking session and an after-party, providing ample opportunities for attendees to connect and mingle.
The shared passion for ComfyUI sparked lively conversations and fostered new connections within the growing community.
This inaugural meetup served as a testament to the vibrant ComfyUI community and the high expectations surrounding the future of AI image generation. As ComfyUI continues to evolve, it promises to unlock new creative possibilities for artists and developers alike.
さらに「New Image with Same Structure」を使って、今度は「Japanese anime style」としてみました。
うーん、ちょっと日本人顔になったけど、もうちょいがんばりたい！ 無料なので。
左上の「Generate another version」を選びます。
右側に追加されていきます。
黒髪とブロンドで悩む私でした。
「New Image with Same Structure」の使い方
同じ構図と構造を維持しながら、周囲の環境を変化させて、コーヒーカップをある景観から別の景観へと移動させてみましょう。この例では、夕暮れの森の中に置かれたコーヒーカップの画像を入力として、このツールを使用して新しい画像を生成します。アシスタントに生成を指示するだけで、 > create a picture of a cup of coffee placed in a forest at sunset (夕暮れの森の中に置かれたコーヒーカップの画像を作成) 次に、ツール設定のプロンプトセクションに文字を入力することで、構造はそのままに、この画像のスタイルと内容を変更することができます。 > a cup of coffee in the desert at sunset (夕暮れの砂漠に置かれたコーヒーカップ)
Today, our CEO, @premakkaraju, announced that legendary filmmaker, technology innovator, and visual effects pioneer, James Cameron, has joined the Stability AI Board of Directors.
Cameron’s addition represents a significant step forward in our mission to transform visual media.… pic.twitter.com/UCirE4WTUs
ポジティブプロンプト: “1boy, solo, young man with afro, male k-pop idol, lovingly looking camera, medium portrait soft light, chop, beautiful model, oval face, vivid”
今回は、「Efficiency Nodes for ComfyUI」というカスタムノードを使用します。「Efficiency Nodes for ComfyUI」は、ComfyUIのワークフローを効率化し、ノードの総数を減らすために開発されたカスタムノードのコレクションです。これにより、画像生成や編集プロセスがよりシンプルかつ効果的になります。
「Save」のプルダウンメニューを開くと、「Save as」と「Save to workflows」が表示されます。「Save as」は、標準の保存機能で、指定した場所にワークフローが保存されます。「Save to workflows」は、保存場所の指定はできず、自動で「ComfyUI/pysssss-workflows」フォルダに指定した名前でワークフローが保存されます。
次に「Load」のプルダウンメニューを開きます。このメニューには、「ComfyUI/pysssss-workflows」フォルダに保存されているワークフローが表示されます（要するに「Save to workflows」で保存したワークフローが表示されるということ）。通常の「Load」では、ファイル選択ダイアログが開き、読み込むワークフローを選択しますが、この機能ではプルダウンメニューに表示されたワークフローをクリックするだけで読み込むことができます。
