SD3生成矢量扁平插画(SD3 generates vector flat illustrations)
5.0
0 reviewsDescription
- 家中常备小工具
- Common Household Tools
- 重要的是矢量
- The important thing is the vector
--------------------------------------------------------------------------------------------
- 比较吃参数,不同的图片细节量不一样,调一个合适的就好,比如LOGO类一套参数,插画类一套参数
- 自制ComfyUI插件节点转svg
- https://github.com/Yanick112/ComfyUI-ToSVG
- 我的出发点是,我想把 ComfyUI 当做一个瑞士军刀来用
- 也可以用SD1.5的模型搭配各种Lora来生成logo平面插画之类的东西再转矢量
- 可批量
- 适合生成插画、LOGO
- 适合平面设计师,UI设计师
- 有问题群里问哈,没细写说明
- 秉持开源精神,不卖课!友好讨论学习
• It depends heavily on the parameters. Different images have varying levels of detail, so just adjust to a suitable level, such as a set of parameters for the LOGO class and a set of parameters for the illustration class.
• Self-made ComfyUI Plugin Node to SVG
• https://github.com/Yanick112/ComfyUI-ToSVG
• My starting point is that I want to use ComfyUI as a Swiss Army knife.
• It can also be used with the SD1.5 model and various Lora to generate logos, flat illustrations, and similar things before converting them to vectors.
• Supports batch processing
• Suitable for generating illustrations, LOGOs
• Ideal for graphic designers, UI designers
• If you have questions, ask in the group; detailed instructions are not provided
• Upholding the spirit of open source, no course selling! Friendly discussion and learning
Node Diagram
Discussion
What is the difference between using the SD3 model versus SD1.5 or even Flux?
Node Details
Primitive Nodes (4)
ConvertRasterToVector (1)
EmptySD3LatentImage (1)
ModelSamplingSD3 (1)
SaveSVG (1)
Custom Nodes (8)
ComfyUI
- VAEDecode (1)
- CLIPTextEncode (2)
- CheckpointLoaderSimple (1)
- SaveImage (1)
- KSampler (1)
- PreviewBridge (1)
- Image Rembg (Remove Background) (1)
Model Details
Checkpoints (1)
SD3\sd3_medium_incl_clips.safetensors
LoRAs (0)