Sign In

五视图角色立绘(Multi-view of people)

827
4.9k
223
Updated: Oct 1, 2023
style
Verified:
SafeTensor
Type
LoRA
Stats
4,878
0
Reviews
Published
Sep 30, 2023
Base Model
SD 1.5
Usage Tips
Clip Skip: 2
Trigger Words
Back View, Back Side View, Front View, Front Side View, Side View, Multi View, Concept Art, Game Character Design
Hash
AutoV2
AA64BE5C30

作者哔哩哔哩个人主页:https://space.bilibili.com/403361177

如果你喜欢这个模型请为我冲杯咖啡吧!

If you like this model, please make me a coffee!

这款lora模型将是最颠覆的一款模型

它不受视角的限制,可生成五视图,三视图,两视图

不受头身比例的限制,可生成八头身,七头身,六头身,五头身,三头身,两头身

兼容不同模型可生成不同的头身比例,

把作者晒图区提供的标准的真人多视图模板下载下来

放入ControlNet中 用open pose固定人物,

将百分百的生成五视图,三视图

模板是我自己用这个模型生成的,我会放在晒图区

This LoRa model will be the most disruptive

It is not limited by viewing angles and can generate five, three, and two views

Not limited by the ratio of head to body, it can generate eight-headed body, seven-headed body, six-headed body, five-headed body, three-headed body, and two-headed body

Compatible with different models to generate different head-to-body ratios,

Download the standard live-action multi-view template provided by the author in the photo area

Put it in ControlNet and fix the character with open pose.

Five views and three views will be 100% generated

The template was generated by myself using this model, and I will put it in the photo area