Sign In

IP DESIGN | 3D可爱化

2.3k
16.0k
329
Verified:
SafeTensor
Type
Checkpoint Trained
Stats
8,458
Reviews
Published
Dec 17, 2023
Base Model
SD 1.5
Trigger Words
3d style
Hash
AutoV2
3C9D7FCA8B
default creator card background decoration
SDXL Training Contest Participant
MG_mango's Avatar
MG_mango

- Model introduction: The large model trained for Pixar style can output the feeling of Pixar 3D, which is similar to the Q version style of Bubble Marte, and can also 3D the 2D picture of the graph, mainly generating a cute wind biased to the Q version (keyword "Meng Meng")

- Creative Overview: The model is a 3D model, trained with mecha mechanical elements characters, and robots, with scenes, etc. In the application, it is more suitable for the pages of the operation design class, and the 2D painting characters can be 3D and avoid the trouble of modeling.

- Generation Suggestion: VAE: cute vae.safetensors Steps:20~30 Recommended: 512*768 (or512*512) Enable the HD Sampler: DDIM

Words: best quality, amazing quality, great quality, normal quality, bad quality, worst quality,

aesthetic tags:, very aesthetic, Aesthetic, displeasing, very displeasing

Welcome to return pictures and comments

V2 is recollected and trained, V2 3D effect is stronger, better applicability

【I discovered a new AI tool called Shakker, one of the best image-to-image tools. You can find it by searching (Shakker) and it helps you:

-Remix: Upload a picture. Just toggle the prompts and you can create stunning images in the same style.

Style conversion :Shakker can not only extract styles, but also switch between various styles.

In addition, Shakker also provides object control, composition control, collage redrawing and more】

-模型简介:针对皮克斯风格训练的大模型,可以输出皮克斯3D的感觉,又类似泡泡玛特的Q版风格,也可以图生图将2D画面3D化,主要生成偏向Q版可爱风(关键词“萌萌“)。

-创意概述:模型为3D化模型,训练有机甲机械元素人物,和机器人,与场景等。在应用上更适用于运营设计类的页面上,可以将2D绘画角色进行3D化而避免了建模的烦恼。

-生成建议:VAE:cute vae.safetensors   Steps:20~30    推荐:512*768(or512*512)建议开启高清修复  Sampler:DDIM

提示词:best quality、 amazing quality 、 great quality 、 normal quality 、 bad quality 、worst quality、

美学标签:、 very aesthetic 、 aesthetic 、 displeasing 、 very displeasing

欢迎各位积极返图和评论

V2 重新收集并训练,V2 的3D效果更强,适用性更好

我发现了一个新的人工智能工具Shakker,它是最好的图像对图像工具之一。你可以通过搜索(Shakker)找到它,它可以帮助你:

-Remix:上传一张图片。只需切换提示,您就可以以相同的风格创建令人惊叹的图像。

风格转换:Shakker不仅可以提取风格,还可以在各种风格之间切换。

此外,Shakker还提供对象控制、构图控制、拼贴重绘等功能。